Improvement of digest2 NEO Classification Code -- utilizing the Astrometry Data Exchange Standard

TL;DR

This paper details enhancements to the digest2 NEO classification software, including updated models, new input formats, and improved accuracy and efficiency in identifying near-Earth objects.

Contribution

The authors integrated the Astrometry Data Exchange Standard (ADES) into digest2, updated the population model, and extended observatory data, significantly improving NEO classification accuracy and computational efficiency.

Findings

Improved NEO identification accuracy.

Reduced computation time with ADES XML input.

Enhanced input format support for better astrometric data handling.

Abstract

We describe enhancements to the digest2 software, a short-arc orbit classifier for heliocentric orbits. Digest2 is primarily used by the Near-Earth Object (NEO) community to flag newly discovered objects for a immediate follow-up and has been a part of NEO discovery process for more than 15 years. We have updated the solar system population model used to weight the digest2 score according to the 2023 catalog of known solar system orbits and extended the list of mean uncertainties for 140 observatory codes. Moreover, we have added Astrometry Data Exchange Standard (ADES) input format support to digest2, which provides additional information for the astrometry, such as positional uncertainties for each detection. The digest2 code was also extended to read the roving observer astrometric format as well as the ability to compute a new parameter from the provided astrometric uncertainties ($RMS'$) that can serve as an indicator of in-tracklet curvature when compared with tracklet's great-circle fit RMS. Comparison with the previous version of digest2 confirmed the improvement in accuracy of NEO identification and found that using ADES XML input significantly reduces the computation time of the digest2.