Final Product Generation: Extended Sources

• Go to the Extended Source Confusion Flag
  
• Go to the LGA Insertion into the 2MASS XSC
  
• Go to the updated table of visually inspected sources.

The final catalog will incorporate the whole sky, roughly twice as large as the 2MASS 2nd Incremental Release. We will use similar methods to generate the final catalog (see Links below), as well as new techniques and processes to undertake this massive job. The 2MASS team has a severe time constraint to pull this off, so we will be working furiously on the catalog to meet the deadline. To this end, we can begin to work on the extended source catalog well before the final pipeline reductions are completed.

Goals: Primary: To generate an extended source catalog that satisfies the Level-1 Science requirements. Secondary: generate a set of catalogs that satisfy less stringent requirements (e.g., reliability).

Data: 2MASS JHKs all sky images, processed with the Version 3 (V3) 2MASS pipeline. Extended sources are identified, characterized and measured using the V3 version of GALWORKS (see Links below). The reduced data include source tables and postage stamp images.

Knowledge: Based on the V3 development tests using RTB and repeated scan data, we know that GALWORKS V3 should produce complete and reliable data for the whole sky (see e.g., V3 Verification of GALWORKS, Notes & Log). There are known problems, however, including artifacts and spurious sources that contaminate the extended source database. We will have to work hard to minimize these bogus sources in the final catalog.

Criteria: The basic criteria for the ext. source catalog are the flux or SNR limit and the star-galaxy discrimination scores. REmoval of dupes, artifacts and problematic sources are also important.

S/N >= 7
E-score or G-score <= 1.4
in-scan and cross-scan duplicates eliminated
known artifacts removed or flagged, depending

Truth File:

A running list of sources checked by hand (or as we say, viewed by "eye"), with a classification scheme as follows:

0 == galaxies
1 == stars
2 == unknown
3 == artifact
4 == double star
5 == galactic fuzz or galactic clusters
6 == triple stars

S/N:

Since GALWORKS computes dozens of integrated fluxes (and hence S/N), we need to decide which S/N to use for cat generation. I propose that we use the following:

S/N (fixed r=7" aperture) >= 7,
or S/N (fiducial elliptical isophotal) >= 7

E/G scores:

G_score <= 1.4,
or
E_score <= 1.4

Duplicate Removal:

Inscan duplicates (overlap from coadds) are removed by TJarrett. Cross-scan duplicates are removed by the exceedingly complicated and byzantine method used by point sources.

The best inscan dupe corresponds to the source most distant from the (inscan) coadd edge.

Artifacts:

Artifacts represent the major reliability contaminent to the XSC. They come in all flavors, including size, color, flux and shape. Take a look at a few representative types here. Most of the effort to produce a clean (reliable) XSC is spent dealing with these nasty false extended sources.

Flagging:

There are many flags to deal with. Too many for the average user. We need to think hard about streamlining our flag situation.

Issue: do we flag or delete sources with known problems from the XSC? (e.g., sources subject to untracked seeing)

Beasty Gallery and Related Issues:

Bright Stars

Meteor Streaks

Edge Flares

Stellar Proximity and False extended sources

Pieces of large galaxies and Galactic fuzz

Other problems

Validation

• Go to the updated table of visually inspected sources.

• Database allsky queries.

See also V3 Verification of GALWORKS, Notes & Log

1. Here is an note from TChester regarding the use of specific queries to uniformly identify artifacts that may then be inspected and (if need be) eliminated from the catalog. Chester_1.note