The Two Micron All Sky Survey (hereafter, 2MASS) is a ground-based, all-sky survey that utilizes the near-infrared band windows of J(1.25 m m), H(1.65 m m) and Ks (2.17 m m). Conceived over a decade ago (cf. Kleinmann et al. 1994), the project has evolved from an extensive prototype engineering phase (cf. Beichman et al. 1998) to the current operational phase in which survey data has been acquired and accumulating since the spring of 1997 (cf. Skrutskie et al. 1997). Two dedicated 1.3-m telescopes, one covering northern declinations and one covering the southern declinations, were designed specifically for 2MASS to provide all-sky uniformity. The data acquisition operations are expected to continue up to 2001 when the sky will be have been covered >98% with satisfactory photometric precision and uniformity. The first public release of 2MASS data will occur in the late fall of 1998, with a large incremental release scheduled for the spring of 1999.

The point source sensitivity limits (10s) are 15.8 (0.8 mJy), 15.1 (1.0 mJy) & 14.3 (1.3 mJy) mag at J,H, Ks, respectively. The extended source sensitivity (10s) is ~1 mag brighter than the point source limits, or 14.7 (2.1 mJy), 13.9 (3.0 mJy) & 13.1 (4.0 mJy) mag at J,H, Ks, respectively. Given the ~2" angular resolution of the image data and the detector sensitivity, the 2MASS survey is well adapted to detection of resolved galaxies out to redshifts >30,000 km/s, as well as compact and diffuse galactic objects. The 2MASS survey is expected to detect over 100 million stars and >1 million galaxies (cf. Chester & Jarrett, 1998). This paper will focus upon the detection, identification and characterization of 2MASS extended sources. Future papers will focus more on specific scientific studies with the 2MASS extended source catalog.

The scientific objectives of the extended-source portion of 2MASS include studies of large scale structure, utilization of the infrared Tully-Fisher relation, a complete survey of the local group of galaxies, and an unprecedented census of galaxies located behind the plane of the Milky Way, often referred to as the "zone of avoidance". As such, survey requirements were established in order to satisfactorily achieve these science goals. In addition to the sensitivity limits given above, the extended source Level-1 Specifications include >90% completeness and 99% reliability for most of the sky (free of stellar confusion). There are no set requirements for observations deep in the Galactic plane.

The level-1 requirements apply to the galaxy catalog derived from the 2MASS database. The catalog reliability criterion is in particular a difficult goal to achieve, necessitating design and implementation of algorithms specifically fabricated to perform star-galaxy separation with 2MASS imaging data, described in section 4. The basic 2MASS data and pipeline reduction overview is given in section 2, including discussion of the point spread function a basic component of star-galaxy separation. In section 3 we describe some of the key parametric measurements made on extended sources and the crucial operational step of background removal. Sections 3 & 4 of this paper describe the algorithms developed to cleanly discriminate between point sources and extended sources. In section 5 we give some examples of the wide array of extended sources that 2MASS is encountering. In section 6 we introduce the extended source catalog and explain how it is generated from the 2MASS database. We also present some basic results from the catalog, including galaxy colors, source counts, completeness and reliability in representative swaths of the sky. We will present more detailed scientific results for field galaxies and galaxy clusters in a future paper.


The first release catalog has several known problems, which are documented in the Caveats file. The general user is strongly urged to read this file. See also the cautionary notes below.

Cautionary Notes:

1. The extended source catalog is just that. The processor tried to identify all sources that were not well-fit by just a single psf. Therefore the database from which the xsc was selected includes objects made out of multiple stars that are close together, such as double and triple stars; artifacts around bright stars due to non-flat background around such bright stars; meteor and plane streaks; as well as true extended objects.


2. We have attempted to reject objects that are not truly extended stars by the following:

    - excluding regions around bright stars;
    - attempting to identify meteor and plane streaks in the database;
    - discriminating against objects made of multiple stars based on a few parameters measured for each source.
The algorithms we have used so far are not perfect, and SOME OF THESE OBJECTS REMAIN. In particular:
    - 2mass cannot obtain accurate magnitude measures for very bright stars due to saturation, and hence the excluded region, which depends on magnitude, is incorrect for many of these very bright stars. In many cases, we will only be able to do an accurate job of identifying such sources when 2mass has completed surveying the entire sky, in order to identify bright sources from other scans which cause artifacts in a given scan.

    - a prototype streak finder in the database relied heavily on finding multiple sources along a line from a single scan. Streaks that produce few sources cannot be found with that algorithm and hence remain in the catalog. Streaks in high source density areas are harder to identify, and hence more probably remain there.

    - some combinations of multiple stars have absolutely identical parameters to those of true galaxies, and hence cannot be excluded based on any current 2mass measurement without also tossing true extended sources.

3. THE XSC IS REALLY TWO CATALOGS IN ONE. In order to get into the catalog, a source in the extended source database first had to pass the filters given in #2 above. Then only spatial extent parameters were used to select sources for entry into the database. The spatial extent parameters used to select sources are combined into an "E" flag by a decision tree algorithm, which attempted to separate true extended sources from sets of multiple stars.

Every source in the xsc passed that decision tree "E" flag criteria. No guarantee is offered to the user as to how many of these sources are truly extended sources as opposed to multiple stars, artifacts, etc. The goal here was just completeness.

The second catalog within the xsc are "galaxy-like" sources. The decision tree used not only spatial extent parameters, but also source colors, to create a "G" flag for each source. Source color is a very powerful tool in order to select galaxies from groups of stars. We attempted to meet stringent reliability standards, which vary as a function of source density, for sources with "G" <= 1.4, henceforth called "G" sources.


4. The "G" sources attempted to meet the following reliability standards: 99% for source density < 103.1 stars per deg^2 brighter than 14th mag, 90% for 103.1 < source density < 103.6, and no reliability standard at all for source density > 103.6.

The goal was to have a high reliability catalog outside the galactic plane. Inside the galactic plane, the goal was shifted to completeness at the expense of reliability. HENCE THE USER SHOULD EXPECT TO FIND ARTIFACTS EVEN AMONG THE "g" SOURCES IN THE GALACTIC PLANE. Thus it clearly is important to check the "density" parameter in order to know the characteristics of the portion of the catalog that you are analyzing.

Spot checks of this first-release catalog show that we have come close to meeting our reliability goals, but probably have not quite achieved it. We will continue to refine our algorithms to try to fully meet those goals by the time of the final reprocessing for the entire 2mass dataset.