Detecting Satellite Breakup Events via Density-Based Spatial Clustering of Space Debris

When artificial satellites disintegrate in orbit due to collisions, explosions, or anti-satellite tests, they generate a large amount of space debris. These debris move in orbit through formation-like flight patterns, which we refer to as debris clusters. Compared to constellation satellites in formation flight, debris clusters have closer mutual distances and higher spatial density. To address this phenomenon, this paper proposes a detection method for satellite breakup events through a two-layer clustering strategy using the DBSCAN algorithm in three-dimensional spatiotemporal domains. The method provides a systematic approach for identifying and characterizing breakup debris through orbital backpropagation and density analysis, serving as a screening tool for potential breakup events. Using the 2009 Cosmos 2251-Iridium 33 collision breakup event as a validation case, experimental results demonstrate that our two-layer clustering approach achieves good clustering accuracy (98.92%) and a positive recall rate (93.20%), with a temporal resolution of 5 s and spatial precision of 21.68 km. The methodology was further applied to analyze the 2024 Intelsat 33E breakup event, successfully identifying sparse debris clusters from Intelsat 33E. The detected breakup time showed an 8 min deviation from the Space-track.org official report.