IV. 2MASS Data Processing

9. Known Asteroid, Comet, Planet, and Satellite Associations

The 2MASS Point Source Catalog contains 15,964 associations with 12,219 known asteroids; 46 associations with predicted positions of 39 known comets; associations with Uranus, Neptune, and Pluto; and 14 associations with planetary satellites. With the exception of planetary satellites, identification of known solar system objects took place as part of standard pipeline data processing.

All-Sky PSC sources that are associated with known solar system objects have mp_flg="1" in their catalog records. The name and orbital data for the objects with which PSC sources are associated are given in the Known Asteroid Detection List, Known Comet Detection List, and Planet and Planetary Satellite Detection List in Section II.6 of this document.

a. Asteroids, Comets and Planets

The strategy used to identify possible detections of known asteroids, comets and planets by 2MASS was to consider the area covered during each scan of a Tile, and the time each point on the sky within the scan was observed. The asteroid and comet ephemerides were then searched to determine which objects may have been within the scan boundaries at the specified epoch. Ephemerides were computed using orbital elements published by the Minor Planet Center for all numbered asteroids, and all multiple-opposition unnumbered asteroids, as well as all periodic comets, and recent nonperiodic comets. The orbits of the planets were included and were taken from the JPL DE403. For the final 2MASS data processing, orbital elements of 80,541 minor planets published by the Minor Planet Center as of 2001 March 9 were correlated with the 2MASS observations. 2MASS observed some objects before their nominal discovery dates (e.g. Sykes et al. 2000 IAUC 7500).

The heliocentric position of the Earth was derived from DE403, and topocentric corrections to the two observing sites were included. Although the ephemeris computations are two-body in nature, the orbits were integrated to dates at 100 day intervals over the Survey period. Therefore, orbits were used that were within ±50 days of the epoch of any 2MASS observation. The ephemeris accuracy was typically 1´´. In addition to the predicted position of the object, the apparent visual magnitude (V_pred)is computed, which can be used to validate proper identification, though a large acceptance window is needed because of unknown lightcurve and color effects, and the line of variation is used to represent the major axis of the error ellipsoid. Distances and phase angle are also computed for purposes of reducing absolute magnitudes to apparent V magnitudes (Bowell et al. 1989, Asteroids II, Binzel et al. eds., Univ. of Arizona Press, pg 524-556).

If a solar system object was predicted to have been within the observed boundaries of a scan during its observation, a search was made of the extracted 2MASS point source lists for objects that positionally correlate with the predicted position. Candidate 2MASS detections are first screened by searching for sources within a coarse window of 30´´ in RA and DEC around the predicted position. For each 2MASS point source within that window, a two-dimensional ² position parameter (the mch_cnf column in the Known Asteroid and Comet and Planet Detection Lists) was computed using the separation between the 2MASS and predicted positions and the combined position error covariance matrix. If the value of the ² is less than 16.0, the association is acceptable (a threshold of 16.0 corresponds to a completeness error of 0.000335; in other words, one correct match out of every 3000 was missed in the attempt to avoid false matches). For example, for a predicted asteroid position uncertainty major axis of 3.0´´ and a minor axis of 1.0´´, this threshold just allows a match with a position discrepancy of 8.5´´ along the major axis and 2.8´´ along the minor axis.

2MASS point sources that have a match ² value below the threshold of 16.0 are flagged as being possible solar system object detections in the PSC with an mp_flg value set to "1". It is otherwise "0". Objects with mp_flg="1" are not positive solar system object identifications, but rather only positional associations. Some percentage of chance associations between predicted asteroid positions and background sources are to be expected, particularly in the Galactic Plane where the background source density is high (see below). Users should examine in detail sources putatively associated with asteroids and comets to determine if the associations are in fact identifications. Information which is useful in determining this includes the 2MASS source color, whether the 2MASS object has a positional association with one of the optical reference catalogs, and ratio of the predicted V magnitude of the solar system object and the associated 2MASS source (e.g [V_pred-K_s]).

Because the precision of 2MASS point source astrometry is typically <0.2´´ with respect to the Tycho 2 reference system, the dominant uncertainty in matching 2MASS candidate sources to asteroids and comets is the uncertainty in orbital predictions. Typical uncertainties are in the range 1´´-5´´, and as expected, the major axis of the asteroid position uncertainty ellipse is generally parallel to the orbital plane. The astrometric precision of 2MASS also means that every sighting of an asteroid or comet can be used to update orbital data for that object. In a few instances, orbital arclengths for an object are unavailable and the prediction position uncertainty ellipse is not calculated. In the case of some sungrazing comets, the orbital arc may be less than one day, in which case the prediction uncertainty ellipse is also not calculated.

1186 asteroids were associated with PSC sources in the First Incremental Data Release PSC, and the properties of the associated objects are discussed by Sykes et al. (2000a Icarus 146, 161-175). 3291 asteroids were associated with PSC sources in the Second Incremental Data Release PSC. 120 of the asteroid associations listed in the First Incremental Release PSC are not included in the Second Incremental Data Release because they fall in the overlap regions between adjacent survey scans that were handled differently in the production of the All-Sky PSC (cf., V.4). 239 of the asteroid associations listed in the Second Incremental Data Release PSC are not included in the All-Sky Release PSC because the scans in which they fell were replaced by scans taken later in the Survey that were deemed to be of higher quality.

Tables 1 and 2 provide a list of the First Incremental Release asteroids that are not found in the Second Incremental Release PSC, and the Second Incremental Release asteroids that are not found in the All-Sky PSC, respectively:

Table 1 - First Incremental Release Asteroids not in the Second Incremental Release.
Table 2 - Second Incremental Release Asteroids not in the All-Sky PSC.

No attempt was made during 2MASS data processing to identify previously unknown minor planets or comets. Such a search might be possible using the repeated observations of the small areas in the overlapping regions in adjacent Tiles.

b. Background Contamination

Because the association of solar system objects and 2MASS sources is based on position only, there is always the chance of association with background astrophysical sources, particularly at low galactic latitudes where the background star density is high. Contamination by chance asteroid associations with reddened background stars in the Galactic Plan can be seen by the redward redward "spike" in the color-color plot of 10,565 PSC sources with asteroid associations and three-band detections shown in Figure 1 (left).

Possible associations with background sources can be identified among the 2MASS asteroid associations several ways:

1. 2MASS sources that are also associated with objects in either the Tycho 2 or USNOA optical catalogs (a="T" or "U"),
2. Sources whose 2MASS positions are relatively distant from the predicted asteroid position (e.g. >>1´´),
3. Sources whose predicted visual magnitudes are fainter than expected based on typical asteroid colors and the 2MASS detection limits.

Figure 1 (right) shows the color-color plot of 3890 asteroid associations in the PSC that have V_pred<17 mag, a separation of less than 1" between predicted position and the coordinates of the associated 2MASS source, and no association with an optical catalog. The "spike" due to reddened Galactic stars disappears. The top panel of Figure 2 shows the V_pred/K_s color-magnitude for 15,923 asteroids associated with sources in the PSC and the 2MASS Survey Point Source Reject Table and Long Exposure (6x) and Calibration Point Source Working Databases (the latter are part of the 2MASS Extended Mission Ancillary Products). Most associations have V_pred-K_s colors scattered about a mean of V_pred-K_s~1.6 mag, which is in the range expected for reflected sunlight. The diagonal locus of points running from lower left to upper right in this diagram is caused by incorrect asteroid/2MASS source associations that have predicted visual magnitudes too faint relative to the associated 2MASS sources. The lower panel in Figure 2 shows the V_pred-K_s distribution with galactic latitude, and illustrates how the chance associations are concentrated towards the Galactic Plane.

Figure 1	Figure 2

c. Planets and Planetary Satellites

Jupiter, Uranus, Neptune, and Pluto were observed during survey operations, and detections of Uranus, Neptune, and Pluto are included in the All-Sky PSC. Uranus was contained in two Tiles in the All Sky Release, so has two entries in the PSC. However, the source is resolved and was not reliably extracted and bandmerged each time it was observed. One PSC entry reports J and H detections and the other reports only a K_s detection. Figure 3 shows J, H and K_s composite Atlas Images of the fields containing Uranus and Neptune; some of the detected satellites of those planets are indicated. The blue color of Uranus and Neptune is caused by strong atmospheric methane absorption which depresses the observed K_s flux.

Jupiter was bright enough that its glare effectively washed out the scan covering it, as well a number of scans adjacent to it. Thus, that region of the sky was reobserved at a later time after Jupiter had moved on, and Jupiter's scan was not included in the All Sky Release.

Planetary satellite associations were made manually after pipeline processing was complete. JPL's Ephemeris Generator softwate was used to determine the predicted positions for all named moons of Mars, Jupiter, Saturn, Uranus and Neptune for all nights of 2MASS observations. A region 60´´ around each position was then searched for matches in the PSC. The radius was chosen to ensure that at least one field object would be detected near the predicted position of the moon (which might not be detected). The resulting tables were then searched for matches between the ephemerides and observation dates. Four moons of Jupiter, five moons of Uranus, and Neptune's moon Triton have listings in the PSC. Ariel, Oberon, Titania and Umbriel are included twice because their location was scanned twice at different times during the Survey. Measurements of Jupiter's satellites detected in the 2MASS scans are presented by Sykes et al. (2000b Icarus, 143, 371).

Figure 3

[Last Update: 2015 November 20, by R. Cutri, B. Nelson, and M. Sykes]