History of Analytical Orbit Modeling in the U.S. Space Surveillance System

Introduction S PACE surveillance became a military mission almost as soon as the first Sputnik satellite was launched on 4 October 1957. In addition to the intense civilian and scientific interest in knowing the locations of space objects, the U.S. Air Force needed a practical way to prevent false missile-warning alarms as satellites transited through the coverage of warning systems, whereas the U.S. Navy needed a way to alert fleet units against possible overhead reconnaissance by satellites. Both needs led to the creation of a complete catalog of detectable space objects, with satellite tracking data forwarded continually to a central processing facility and updated orbital data distributed routinely to defense users. Naturally, the catalog also served, and still serves, a variety of civilian and scientific purposes. To date, cataloged orbits have been represented by some type of mean orbital elements, although the operational models have become more elaborate over time as computers have improved. It has always been known that special perturbations can provide better accuracy than general perturbations, at least in principle. However, the sheer number of satellite orbits to be processed for the catalog has meant that only simplified, analytic orbit models could be used in practice. Only recently, with the advent of multiprocessor computer techniques, has it been possible to consider maintaining the satellite catalog with special perturbations,1 and this implementation is in progress. Throughout the history of orbital mechanics, the interaction between the development of orbit models and the development of computational facilities has been often noted but seldom studied. Although we cannot offer such a study here, we can note that this interaction has been crucial in the development of U.S. space surveillance capabilities in general and in the development of the basic orbit models in particular.