S318p.07 — On the degradation of asteroid astrometry due to background objects

Date & Time

Aug 4th at 6:00 PM until 7:30 PM




Rating ( votes)

Author(s): Anatoliy Ivantsov2, Siegfried Eggl1, Daniel Hestroffer1, William Thuillot1, Marcelo Assafin3

Institution(s): 1. Paris Observatory, 2. Technion - Israel Institute of Technology, 3. UFRJ Observatorio do Valongo

The fact that images where an asteroid is apparently close to a star or another background object can cause a potential degradation of astrometric accuracy has received little attention so far. The reason for this bias lies in the non-zero background gradient that will change the brightness distribution in the image of the asteroid. This leads to an apparent shift in its position in the direction of the nearby object. In the worst case reduction software suites can no longer differentiate between the sources representing the asteroid and the background star which results in a complete misidentification. Since today many astrometric measurements are conducted on an automated basis, such a problem has to be taken seriously. While there is a possibility for observers to identify these problematic measurements using the observational notes by the IAU Minor Planet Centre (MPC) in the current observational format, this option has not been used efficiently. In fact, we have conducted a search for problematic configurations in all asteroid positions observed during the last 30 years that are listed in the MPC database. To this end we have cataloged stars and non-stellar objects from the PPMXL, UCAC4, NOMAD, USNO-B1.0, SDSS catalogs within 8″ of all astrometric measurements of asteroids. We have found more than 2 million cases within 3.5″ where astrometric measurements should have been compromised. Using statistics on residuals from observations of the MPC, and some assumptions, we have identified those observations which may contain the aforementioned bias. The elimination of this error is possible by jointly modeling the images of the asteroid and the background stars present in the measuring aperture. We provide an expression for the correction of this error using additional information about the measurement details. We recommend that images containing background sources close to the asteroid's position should either receive lower weights in the orbital fitting process, or be eliminated entirely. Such a procedure can be shown to improve the orbits of near-Earth asteroids, and it can facilitate the dynamical linking of the newly discovered objects.