Time series prediction for satellite ballistic coefficients

ABSTRACT: The leading errors in computing future positions of satellites in Low Earth Orbit (LEO) are generally due to inaccuracies in the thermospheric density and the prediction thereof. The use of Dynamic Calibration Atmosphere (DCA) has significantly reduced these traditional sources of error and variations seen in ballistic coefficients can now be attributed to unmodeled satellite frontal area changes. When the orbit of a satellite needs to be predicted, there is no way of knowing the correct value of the ballistic coefficient for the orbit prediction interval thus a value is assumed. The assumed value of the ballistic coefficient will cause the predicted orbit to be in error. Hence, considerable improvement in the quality of orbit prediction can be achieved by reducing the error in the assumed value of the ballistic coefficient. The ballistic coefficient for prediction is usually obtained from the estimated value prior to the prediction. Instead of assuming the previous estimated value, an analysis of the time series of a history of the estimated values may reveal some characteristics which then can be used to minimize prediction error. BENEFIT: Many Air Force and other agencies use satellite prediction products produced in the JSpOC. Therefore, the potential for greatly improved accuracy of the space catalog has far reaching applicability to a wide range of DoD and commercial users.