Abstract
Predicting the re-entry of space objects enables the risk they pose to the ground population to be managed. The more accurate the re-entry forecast, the more cost-efficient risk mitigation measures can be put in place. However, at present, the only publicly available ephemerides (two line element sets, TLEs) should not be used for accurate re-entry prediction directly. They may contain erroneous state vectors, which need to be filtered out. Also, the object’s physical parameters (ballistic and solar radiation pressure coefficients) need to be estimated to enable accurate propagation. These estimates are only valid between events that change object’s physical properties, e.g. collisions and fragmentations. Thus, these events need to be identified amongst the TLEs. This paper presents the TLE analysis methodology, which enables outlying TLEs and space events to be identified. It is then demonstrated how various TLE filtering stages improve the accuracy of the TLE-based re-entry prediction.