Keys, Sonia, Vereš, Peter, Payne, Matthew J., Holman, Matthew J., Jedicke, Robert, Williams, Gareth V., Spahr, Tim, Asher, David J., and Hergenrother, Carl. 2019. "The digest2 NEO Classification Code." Publications of the Astronomical Society of the Pacific, 131 064501. https://doi.org/10.1088/1538-3873/ab1157.
We describe the {digest}2 software package, a fast, short-arc orbit classifier for small solar system bodies. The {digest}2 algorithm has been serving the community for more than 13 yr. The code provides a score, {D}2, which represents a pseudo-probability that a tracklet belongs to a given solar system orbit type. {digest}2 is primarily used as a classifier for Near-Earth Object (NEO) candidates, to identify those to be prioritized for follow-up observation. We describe the historical development of {digest}2 and demonstrate its use on real and synthetic data. We find that {digest}2 can accurately and precisely distinguish NEOs from non-NEOs. At the time of detection, 14% of NEO tracklets and 98.5% of non-NEOs tracklets have {D}2 below the critical value of {D}2=65.94 % of our simulated NEOs achieved the maximum {D}2=100 % and 99.6% of NEOs achieved {D}2≥slant 65 at least once during the simulated 10-year timeframe. We demonstrate that {D}2 varies as a function of time, rate of motion, magnitude and sky-plane location, and show that NEOs tend to have lower {D}2 at low Solar elongations close to the ecliptic. We use our findings to recommend future development directions for the {digest}2 code.