Several fields are selected to test final Catalog Generation algorithms, implementation and analysis.
Name | ra0 (deg) | dec0 (deg) | N_dec_bands | Tiles |
South1 | 75.0 | -27.0 | 3 | 309418-309458,312218-312256,314900-314937 |
Equator1 | 90.0 | +12.0 | 2 | 6665-6685,203721-203740 |
Equator2 | 283.3 | +6.0 | 2 | 202340-202362,205306-205328 |
Polar1 | - | 90.0 | 1 | 29514-29824 |
0/0 | 0.0 | 0.0 |
0/0
Field centered on 0h meridian and celestial equator to
test coordinate calculations.
South1 | Equator1 | Equator2 | Polar1 | 0/0 | |
Number of Scans: | 118 | 41 | 46 | 311 | |
Number of WDB Point Sources: | 1205906 | 957267 | 3303690 | 3788450 | |
Number of CatGen Point Sources: | 250732 | 546030 | 2912796 | 848192 | |
Number of WDB Extended Sources: | 4498 | 1810 | 121 | 10217 | |
Number of CatGen Extended Sources: | 4135 | 1467 | 94 | 8589 | |
List of Scans in Field | X | X | X | X | |
Map of Ks<13m Sources in Galactic Coords (red - R1-saturated in >1 bands; blue - R1-saturated in 3-bands) | X | X | X | X |
South1 | Equator1 | Equator2 | Polar1 | 0/0 | |
Drawn from Working DB (available in GATOR and rmt_grit) | X | X | X | X | |
Multiple Source Processing Complete (use_src, dup_src, spos and sdet values set) | X | X | X | X | |
DB_MAPCOR Processing Complete (prg_flg and cc_flg updated) | X | X | X | X | |
Extended Source Processing Complete (cc_flg and vc values set) | X | X | X | X | |
Proximity Analysis Complete (prox and pxpa values set) | |||||
"Catalog" field set | X | X | X | X |
I have attached a plot that shows the distribution in nearest neighbor of sources in the test fields (only sources with use_src=1 were used) where the nearest neighbor was in a different scan. The dup resolution algorithm should have killed all neighbors in a different scan closer than 2" and indeed it has without exception. I think I understand why the distribution turns over at larger separations but do others agree?
I have attached two figure that may be useful a discussion of how sources with use_src=0 and dup_src=0 (i.e. sources that are identified as not having a matching source in an overlap region where that is geometrically possible) should be treated in the catalog selection process. Sherry has indicated that there is a bug in the dup resolution software and the plots are made with data processed with the bug. However, I don't think the bug affects many sources. I propose that sources not be rejected from the faint source list because use_src=0 and dup-src=0 and that similar sources in the catalog region of snr either be accepted as well or be placed in the faint source list. RaeGhost-Kill Fraction vs. SNR (3/31/02)
A bug in the duplicate source resolution code was found that was inadvertantly dropping some non-duplicate sources that should be kept for the catalog. That bug was fixed and the duplicate resolution rerun. The following plots show the ghost kill-fractions after the bug fix. Rae reports: "It seems that the fix recovered a few more sources in the catalog region than in the faint source region."
Ghost-Kill Fraction vs. SNR (4/3/02)
Ghost-Kill Fraction vs. SNR - expanded scale (4/3/02)
Attached is a whirlgif which now shows the (very nice) color-color/mag diagrams for the first cut query. I'm blinking between the entire selected set against a kludge in which I prevent the query from selecting the snr>5 in any one band faint sources. You can see they fill in rather nicely.
I was curious about the bright spot in John Carpenter's plots near the north equatorial pole. Since the galaxy distribution is independent of the star distribution I was curious about whether or not the galaxies would show a bright spot as well.NCP Dup-Resolved Galaxy Gatalog
(M. Skrutskie - 4/1/02) John notes that the bright spot is NGC 188. I've attached the coadd image. Since his is a star counts rather than a flux map the cluster really stands out.
Last Updated: 13 May 2002
R. Cutri - IPAC