Advanced techniques for the stochastic fragmentation event characterization from a single fragment orbital state

The growing population of objects in low Earth orbits poses increasing risks to satellites. Fragmentation events—collisions, explosions, and material degradation—are the main contributors to orbital overcrowding, generating further debris. This research develops a tool for characterizing fragmentation events, with emphasis on determining their epochs. Knowing the epoch improves observation planning and debris prediction. This tool also achieves fragment–event association and the rejection of unrelated objects. This enhances the cataloguing of newly generated debris. The method models the event as a collision between the parent body and a fragment, computing collision probabilities over time. The highest probability indicates potential fragmentation epochs, from which a single estimate is derived. Gaussian mixtures are used to represent and propagate uncertainty, offering more accurate probability computations by capturing non-linearities that Gaussian assumptions could not. Validation through numerical analysis shows that, while perturbations and orbit determination errors degrade accuracy, the correct epoch is always among the candidates. Real data tests confirm the method robustness in operational scenarios. The algorithm proves reliable and requires only one fragment state and the parent ephemeris, making it highly practical.