A systematic method has been developed for managing long-term collision risk posed to operational satellites in geosynchronous orbit. Long-term collision risk reduction is achieved by proper selection of local collision probability thresholds that trigger actions to reduce risk. Such actions can be requests for more accurate orbital data, special sensor tasking, or collision avoidance maneuvers. The cost of collision risk reduction is measured by the frequency of actions taken to reduce the collision risk. This action frequency is dependent on the accuracy of the orbital data and the frequency of close approaches. A case study was performed for a set of satellites. The analysis used position error models for two-line element sets. A conjunction simulation was used to process approximately three years of archived orbital data, including publicly available two-line element sets, in order to generate conjunction statistics. From these results, a graphical representation called a chi-plot was generated. This plot permits the selection of thresholds as a function of total risk reduction and tolerable action frequency. Results of the study indicate that collision risk management in geosynchronous orbit can be very costly using data of insufficiently high accuracy, because data errors induce high action frequency for even modest amounts of collision risk reduction. (C) 2004 COSPAR. Published by Elsevier Ltd. All rights reserved.
