Meteoroid stream identification with HDBSCAN unsupervised clustering algorithm

Eloy Peña-Asensio, Fabio Ferrari · 2025

Read Meteoroid stream identification with HDBSCAN unsupervised clustering algorithm by Eloy Peña-Asensio, Fabio Ferrari (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-01501.txt.

Description

Meteor showers come from streams of debris, and sorting meteors into their streams reveals where the debris came from. Eloy Peña-Asensio and Fabio Ferrari test an unsupervised clustering method, HDBSCAN, on the CAMS meteoroid orbit database. They compare its groupings with the standard CAMS look-up table, using three feature sets built from geocentric measures and orbital elements. A matching algorithm pairs the new clusters with known streams, and standard scores judge the agreement.

With the geocentric feature set, the method confirms 39 streams, 21 of them strongly matching the catalog. Orbital features give 30 streams with fewer strong matches, and weak showers remain hard to detect. One cluster selection style, called eom, beats the other throughout. The authors conclude that the clustering approach is statistically more coherent than the look-up table, though physical checks are still needed, and the minimum cluster size must be chosen with care.

Catalog record

CREATOREloy Peña-Asensio, Fabio Ferrari
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE69 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-01501

Browse the full library · About the Tower of Babel