Statistical equivalence of metrics for meteor dynamical association

• Published in: Advances in space research Vol. 74; no. 2; pp. 1073 - 1089
• Main Authors: Peña-Asensio, Eloy, Sánchez-Lozano, Juan Miguel
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.07.2024
• Subjects: Meteor; Meteor showers; Meteoroid streams; Meteoroids; Statistics; Meteoroid streams; Meteor showers; Meteoroids; Statistics; Meteor
• ISSN: 0273-1177; 1879-1948
• DOI: 10.1016/j.asr.2024.05.005

•Evaluated the performance of various dissimilarity criteria and distance metrics for meteor association with CAMS database, as well as computed their optimal thresholds.•The sEuclidean metric excels in meteor shower association and identifying sporadic meteors, with DSH and ϱ2 closely behind, the latter showing the closest equivalence to Machine Learning metrics.•Machine Learning distance metrics demonstrate their potential to match or even surpass tailored orbital similarity criteria in meteor dynamical associations. Meteor showers, originating as a result of the activity of comets or the disruption of large objects, provide a unique window into the composition and dynamics of our Solar System. While modern meteor detection networks have amassed extensive data, distinguishing sporadic meteors from those belonging to specific meteor showers remains challenging. In this study, we statistically evaluate and compare four orbital similarity criteria within five-dimensional parameter space (DSH,DD,DH, and ϱ2) to study dynamical associations using the already classified meteors (manually by a human) in CAMS database as a benchmark. In addition, we assess various distance metrics typically used in Machine Learning with two different vectors: ORBIT, grounded in heliocentric orbital elements, and GEO, predicated on geocentric observational parameters. To estimate their degree of correlation and efficacy, the Kendall rank correlation coefficient and the Top-k accuracy are employed. The statistical equivalence of the Top-1 results is examined using the Kolmogorov–Smirnov test and the percentage of Top-1 agreement is calculated on an event-by-event basis. Additionally, we compute the optimal cut-offs for all methods for distinguishing sporadic background events. Our findings demonstrate the superior performance of the sEuclidean metric in conjunction with the GEO vector. Within the scope of D-criteria, DSH emerged as the preeminent metric, closely followed by ϱ2. The Bray-Curtis metric displayed an advantage compared to the other distance metrics when paired with the ORBIT vector for Top-k accuracy, however, the Cityblock metric is more effective when considering the sporadic background. ϱ2 stands out as the most equivalence to the distance metrics when utilizing the GEO vector and the most compatible with GEO and ORBIT simultaneously, whereas DD aligns more closely when using the ORBIT vector. The stark contrast in DD’s behavior compared to other D-criteria highlights potential inequivalence. Our results suggest that geocentric features provide a more robust basis than orbital elements for meteor dynamical association. Most distance metrics associated with the GEO vector surpass the D-criteria when differentiating the meteoroid background. Accuracy displayed a dependence on solar longitude with a pronounced decrease around 180° matching an apparent increase in the meteoroid background activity, tentatively associated with the transition from the Perseids to the Orionids. Considering lately identified meteor showers, ∼27% of meteors in CAMS would have different associations. This work unveils that Machine Learning distance metrics can rival or even exceed the performance of tailored orbital similarity criteria for meteor dynamical association.