Skip to main content
SearchLoginLogin or Signup

Digest2 - ADES-astrometry enhanced orbit classification code

Presentation #404.03 in the session Asteroids: Observational Surveys (Poster)

Published onOct 23, 2023
Digest2 - ADES-astrometry enhanced orbit classification code

We present an updated version of the digest2 short-arc orbit classification code, capable of reading astrometry in the Astrometry Data Exchange Standard (ADES) format, which allows for per-detection user-provided astrometric uncertainty. Digest2 has been widely used by the Near-Earth Object (NEO) community for flagging newly discovered objects and has been a part of the NEO discovery process for over 15 years. In this new version, we have made several enhancements. Firstly, we have updated the solar system population model used to weight the digest2 score based on the current 2023 catalog of known solar system orbits. Additionally, we have expanded the list of mean astrometric uncertainties from 35 to 143 observatory codes. Moreover, the improved code now includes the ability to process roving observer astrometry. We have also introduced a new parameter, denoted as RMS’, which is computed from the provided astrometric uncertainties. This parameter serves as an indicator of in-tracklet curvature when compared with the tracklet’s great-circle fit RMS, which would suggest motion closer to Earth. To handle user-submitted astrometric values that are either grossly under- or overestimated, we have implemented a floor and ceiling for acceptable uncertainties. To evaluate the performance of the enhanced code, we compared it to the old code with its previous configuration using four distinct and large data samples. These samples consisted of millions of tracklets from known objects such as main-belt objects representing the background population, NEOs, Near-Earth Object Confirmation Page (NEOCP) candidates, and unknown objects from the isolated tracklet file (ITF). The results indicate that the updated code reduces false-positive NEO identifications and exhibits significantly improved speed when reading the ADES input. The RMS’ parameter successfully flagged several NEO tracklets showing potential curvature. However, it also identified nearly 2000 tracklets of main-belt objects as false-positive curvatures. Further analysis of the flagged main-belt objects revealed that the false-positive results were caused by underestimated reported astrometric uncertainties, which were much smaller than the mean astrometric error for the given observatory code. In addition, we discussed a new method of assessing in-tracklet curvature by computing along and cross-tracklet acceleration when per-detection uncertainties are provided. We have been analyzing low NEO-digest score Pan-STARRS tracklets for a few months and have identified a few candidates worthy of further investigation.

No comments here