Since the first discovery of a Trans-Neptunian object (TNO) a few years ago (Jewitt & Luu 1993), the search for other similar objects has involved more and more observers and telescopes. Nowadays about 60 TNOs are officialy registered by the IAU and one can expect many other discoveries in the near future. This observational effort involves more and more astronomical data that need data processing in order to identify, usually near the sky background noise, the very faint TNOs.
The usual procedure used to detect new TNOs has been described by Jewitt & Luu (1988). It consists in obtaining several images of a same field of view, separated by about 1 or 2 hours (3 images is the minimum). These images are examined in order to detect all the moving objects and to measure their velocity, which is the main criterion used to classify the objects as TNOs or asteroids. Indeed Kepler's third law shows that the apparent velocity (expressed, for instance, in arcsecond/hr) of an object like a TNO (with an orbit supposed nearly circular) is directly related to its heliocentric distance. For an object observed near opposition, as it is usually done, we have: , R being the heliocentric distance of the observed TNO (expressed in AU).
For a typical value R=40 AU one obtains arcsec/hr. Indeed, with three or four images obtained with an integration time of about 15 mn and separated by about two hours, a TNO appears as a point-like object in different places separated by a few arcseconds. Thus the detection criteria for TNOs is different than that for asteroids; the main difference being that in a single image a TNO appears as a point-like object rather than as a streak.
So far several authors have published results obtained with automatic detection procedures. First, Levison & Duncan (1990) described the results obtained with a search program dedicated to TNOs. This program did not manage to detect any TNOs, but the authors mention the use of their own program that is based on an object-list algorithm. More recently Irwin et al. (1995) were the first to detect TNOs with an automated algorithm. This algorithm was also especially designed for the search for TNOs, but used a different method (creation of difference maps, assuming that the seeing does not vary too much for the different images). Also, Trujillo & Jewitt (1998) described an automatic algorithm, similar to the one described by Levison & Duncan (1990), which was used to process some data obtained with the UH 8k CCD mosaic Camera.
In this paper we describe a new program especially optimized for the search of new TNOs with a "standard" set of images. This program has the key advantage, when compared to the works mentioned above, of being designed to be used with one of the main astronomical data processing packages (MIDAS). This feature greatly facilitates the processing of new images.
© European Southern Observatory (ESO) 1999
Online publication: August 13, 199