Internal Completeness Analysis for Catalog Point Sources
Found in Two-Scan Coverage Areas (SPOS=2)
Found in Two-Scan Coverage Areas (SPOS=2)
This analysis examines the internal completeness of 2MASS Final Catalog point sources. A separate analysis exists for the Final Catalog extended sources.
The Level 1 requirements for point source completeness in absence of confusion is 99% at 15.8,15.1,14.3 magnitudes for J,H,K, respectively.
Source Selection
Sources selected for this analysis came from non-confused regions of the sky; fell in the two-scan only overlap regions and did not have magnitude set by band fill [PRG_FLG=FIL] or upper limit [PRG_FLG=UPL]. Each band [J,H,K] was examined independently.
- Select non-confused sources:
- Outside the Galactic Plane, abs(GLAT) > 30 degrees.
- DUP_SRC <= 1.*
- Select sources with RD_FLG[x} == [1,2,3]. This removed a source having band magnitude set by band fill [FIL] or upper limit [UPL].
- Select sources with SPOS = 2 , where SPOS is the number of possible detections for a source (m). This assures the source fell within a two-scan coverage area.
* ((DUP_SRC <= 1 and USE_SRC = 1) or (DUP_SRC = 0 and USE_SRC = 0)).
Processing
Two histogram files were made for each band [J,H,K] from 8 to 20 magnitude using a bin size of .5 magnitude.
One contained sources with SDET = 2, where SDET is the number of actual detections for a source (n), and the other contained sources where SDET=1.
Completeness was computed for each bin and each band using equation {1}.
Results were plotted. Lines on the plots show the Level 1 magnitudes and completeness requirements. Points on the completeness plots are at the mid-bin position.
The completeness algorithm
where
sdet2 = number of sources having an apparition in each overlapping scan.
sdet1 = number of sources having an apparition in one of two scans.
Results
The results show completeness between the .99 (Level 1 requirement) and 1.0 in all bands out to the Level 1 cut off magnitudes of 15.1, 15.8 and 14.3 in J, H and K, respectively. In examining some of the sources with SDET = 1 (n = 1 and m = 2) and magnitudes between 10 and 12, it appears that inconsistencies in the distance-to-edge computation may account for some of the less than perfect completeness results. Computation of the distance-to-edge in the duplicate resolution process were made utilizing a more accurate estimate of the scan corners than used in the same computation for the point source database. This resulted in some sources near the 10" scan edge threshold getting an SPOS of 2 while the matching apparition fell outside the 10" threshold and was not selected. By pre-selecting sources away from any scan edge, the completeness numbers should improve. The extent of the improvement is TBD. Plots of the completeness follow:
Plots of the completeness results:
|
|
|
| Figure 1 | Figure 2 | Figure 3 |
Tables of the completeness results:
Additional analysis of the overlap regions showing delta mags, delta RA and delta Dec can be found here
[Last Updated: 2003 Jan 22; by Sherry Wheelock]