A Trade Study of Thermosphere Empirical Neutral Density Models

Abstract:

Accurate orbit prediction of space objects critically relies on modeling of thermospheric neutral density that determines drag force. In a trade study we have investigated a methodology to assess performances of neutral density models in predicting orbit against a baseline orbit trajectory. We use a metric defined as along-track error in a day a satellite is predicted to have for a given neutral density model when compared to its GPS positions. A set of ground truth data including Gravity Recovery and Climate Experiment (GRACE) accelerometer and GPS data, solar radio F10.7 proxy and magnetic activity measurements are used to calculate the baseline orbit. This approach is applied to compare the daily along-track errors among HASDM, JB08, MSISE-00 and DTM-2012 neutral density models. The dynamically calibrated HASDM model yields a daily along-track error close to the baseline error and lower than the other empirical models. Among the three empirical models (JB08, MSISE-00 and DTM-2012) the MSISE-00 model has produced the smallest daily along-track error. The results suggest that the developed metric and methodology could be used to assess overall errors in orbit prediction expected from empirical density models. They have also been adapted in an analysis tool Satellite Orbital Drag Error Estimator (SODEE) to estimate orbit prediction errors.