Speaker
Mr
Aleksander Lidtke
(University of Southampton)
Description
Upper stages of rockets are large objects, which contain components that are known to be able to survive atmospheric re-entry. Such surviving material, for example propellant tanks, will impact Earth’s surface and might cause ground casualties. Predicting the satellite re-entry, and thus also impact location is notoriously difficult; re-entry prediction is associated with uncertainty in the order of 10% of the remaining lifetime in-orbit. This makes managing the ground casualty risk, by issuing actionable impact warnings, challenging. Thus, the risk posed by spacecraft re-entries will be reduced if the accuracy with which such events can be predicted is improved.
At present, Two Line Element sets (TLEs) are the only publicly available data that can be used for re-entry prediction of a space object. However, there is a number of factors that, if unaddressed, could reduce the accuracy of re-entry prediction based on TLEs:
1. The quality of TLEs of an object is not homogeneous; sometimes TLEs of low quality or even belonging to a different object are published.
3. Occasionally, the object or its orbit can be altered by collisions, fragmentations or space weather phenomena. Such space events render the TLEs of the object from before the event inapplicable to its new, changed state.
4. TLEs do not provide information on space object parameters, such as ballistic coefficient (BC) or solar radiation pressure coefficient (SRPC). TLEs only include the B* parameter that accounts for combined atmospheric drag and solar radiation pressure forces, not BC and SRPC individually.
5. TLEs can only be propagated using the SGP4/SDP4 propagator. However, this propagator is based on the Brouwer theory and, therefore, only models the largest perturbations affecting a satellite. The many assumptions of the theory can severely limit the accuracy of the resulting propagation and thus of the re-entry prediction.
6. TLEs are not supplied with uncertainty information, e.g. a covariance matrix. It is thus challenging to estimate the accuracy with which the re-entry is predicted based on these ephemerides.
In order to overcome these difficulties in TLE-based re-entry prediction, a multi-step procedure is proposed. The first step consists of analysing TLEs, with the goal of identifying outliers, space events, and dynamical phases of re-entry, where the drag is relatively high or low. The filtered TLEs are then used to estimate the unknown spacecraft BC and SRPC. The last step consists of performing an orbit determination in which the TLEs (and derived osculating elements) are used as pseudo-observations.
This paper presents the approach adopted to process the TLEs to improve the accuracy of re-entry prediction. This processing is based on methods previously employed to detect space weather events, which slide a window through the orbital elements contained in the TLEs or derived quantities. Details of the algorithm, which enables TLEs of varying quality and generated in different phases of re-entry to be analysed using the same method, are given. Then, the method used to distinguish between space events and outlying TLEs is described. The trade-off between the number of false positives and negatives, i.e. incorrectly identified and missed outliers, is emphasised. The results of applying the TLE processing methodology to several example rocket bodies are presented in detail and discussed in the context of the accuracy of the resulting re-entry prediction.
| Applicant type | First author |
|---|
Primary author
Mr
Aleksander Lidtke
(University of Southampton)
Co-authors
Dr
Camilla Colombo
(University of Sotuhampton)
Mr
David Gondelach
(University of Southampton)
Dr
Hugh Lewis
(University of Southampton)
Mr
Quirin Funke
(European Space Operations Centre)
Dr
Roberto Armellin
(Universidad de La Rioja)
Dr
Tim Flohrer
(European Space Operations Centre)
