Alejandro Pastor, GMV; Jan Siminski, ESA/ESOC Space Debris Office; Guillermo Escribano, University Carlos III of Madrid; Manuel Sanjurjo-Rivo, University Carlos III of Madrid; Diego Escobar, GMV
Keywords: fragmentation, detection, tracking, break-up, cataloging, association, correlation
Abstract:
Break-up events represent the dominant source of objects in space catalogs, surpassing half of the total population. The number of such events includes explosions, collisions or anomalous events resulting in fragmentation and is estimated to be higher than 630. The contribution of each event to the population is complex and diverse. Two of the most massive events, involving a number of fragments in the order of the thousand are the Fengyun 1C anti-satellite weapon test in 2007 and the accidental collision of Cosmos 2251 and Iridium 33 in 2009, accounting for over 30% of all cataloged space objects. In 2021, three main break-up events happened: the failure of NOAA 17 (10th March), an accidental collision of YunHai 1-02 with a small mission-related debris object (18th March) and the destruction of Cosmos 1408 in an anti-satellite weapon test (15th November). As of today, the number of detected and cataloged fragments by the 18th Space Control Squadron (SPCS) associated to these events is 115 (1 decayed), 37 (4 decayed) and 1561 (243 decayed), respectively.
The early detection of the fragments generated during these not so uncommon events, almost four per year on average over the last decade, poses a complex challenge for space objects catalog build-up and maintenance processes. These fragments are first a dense cloud of debris, making the identification of individual objects rather difficult. Then, a trade-off between detection time and reliability arises, where time favors the spreading of the objects along the orbit, thus reducing the probability of false associations and the uncertainty of the estimated trajectories, since more data is available. However, the provision of Space Situational Awareness products and services during the few first days after a break-up event can be crucial to avoid collisions between the fragments and other space objects, particularly in highly congested regimes, such as Low Earth Orbit. In this regard, reducing the time required to establish the trajectories of the fragments may enable the execution of collision avoidance maneuvers, in the case of operational space objects with maneuver capabilities, and analyze potential collisional cascade events, which may endanger the space environment. The evolution of the cataloging process of the fragments from Cosmos 1408 is a clear example of this complexity: 185 fragments detected and cataloged two weeks after the event (1st December), 718 the next month (903 total as of 1st January) and 494 the next month (1397 total as of 1st February).
This paper studies the whole cataloging process after a break-up event, starting from a catalog with no fragments from the fragmentation under-analysis, and until a well-established orbit is obtained for all the fragments, using a ground-based sensor network. First, the observations enter a multi-sensor multi-target track-to-track association algorithm in charge of grouping observations belonging to the same objects. To resolve the ambiguity, particularly shortly after the event, hypotheses about tracks belonging to the same fragment are generated, scored, pruned, and promoted, only when there is enough confidence, leading to the initialization of new objects in the catalog. Special emphasis is placed on the figure of merit used for the association, considering both covariance and residuals-based approaches. As soon as the catalog is populated, a track-to-orbit association algorithm is responsible for the correlation of observations and already cataloged orbits. This alleviates the track-to-track association and enables the update of the orbital estimates, required for maintaining the catalog. The use of dynamical models of varying fidelity, including analytical, semi-analytical and numerical propagators, during the track-to-track association, track-to-orbit correlation and orbit determination processes is investigated as a way to enable a real-time capability while not jeopardize the accuracy of the final products.
To do so, a fragmentation event has been simulated with NASA Standard Breakup Model, matching the available public information about the Cosmos 1408 break-up event. The trajectories of fragments are obtained via high fidelity numerical propagation. Besides, observations from a ground-based sensor network are simulated, including also typical sensor measurement accuracies. Then, the simulated data is provided to the proposed framework and the performance is analyzed on a time basis. This includes the confusion matrix evolution (true positives, false positives, and false negatives) during both the track-to-track and track-to-orbit association, as well as the accuracy of the estimated trajectories of the fragments. The considered metrics aim at evaluating the robustness and efficiency of the framework, conceived for real operational environments. In this sense, the simulated scenario is used to improve knowledge and understanding of the break-up events, so that the time required to identify and track the fragments generated on break-up events can be reduced. Results include the distributions of the figure of merit of the hypotheses during the association processes, as a function of orbital elements, association time and time since the break-up, among others. Besides, the temporal evolution of the accuracy of the cataloged orbits is evaluated and discussed, as well as their corresponding uncertainty.
Date of Conference: September 27-20, 2022
Track: Space Debris