ATMOSPHERIC DENSITY ESTIMATION USING SATELLITE PRECISION ORBIT EPHEMERIDES

Abstract

The current atmospheric density models are not capable enough to accurately model the atmospheric density, which varies continuously in the upper atmosphere mainly due to the changes in solar and geomagnetic activity. Inaccurate atmospheric modeling results in erroneous density values that are not accurate enough to calculate the drag estimates acting on a satellite, thus leading to errors in the prediction of satellite orbits. This research utilized precision orbit ephemerides (POE) data from satellites in an orbit determination process to make corrections to existing atmospheric models, thus resulting in improved density estimates. The work done in this research made corrections to the Jacchia family atmospheric models and Mass Spectrometer Incoherent Scatter (MSIS) family atmospheric models using POE data from the Ice, Cloud and Land Elevation Satellite (ICESat) and the Terra Synthetic Aperture Radar - X Band (TerraSAR-X) satellite. The POE data obtained from these satellites was used in an orbit determination scheme which performs a sequential filter/smoother process to the measurements and generates corrections to the atmospheric models to estimate density. This research considered several days from the year 2001 to 2008 encompassing all levels of solar and geomagnetic activity. Density and ballistic coefficient half-lives with values of 1.8, 18, and 180 minutes were used in this research to observe the effect of these half-life combinations on density estimates. This research also examined the consistency of densities derived from the accelerometers of the Challenging Mini Satellite Payload (CHAMP) and Gravity Recovery and Climate Experiment (GRACE) satellites by Eric Sutton, from the University of Colorado. The accelerometer densities derived by Sutton were compared with those derived by Sean Bruinsma from CNES, Department of Terrestrial and Planetary Geodesy, France. The Sutton densities proved to be nearly identical to the Bruinsma densities for all the cases considered in this research, thus suggesting that Sutton densities can be used as a substitute for Bruinsma densities in validating the POE density estimates for future work. Density estimates were found using the ICESat and TerraSAR-X POE data by generating corrections to the CIRA-72 and NRLMSISE-00 atmospheric density models. The ICESat and TerraSAR-X POE density estimates obtained were examined and studied by comparing them with the density estimates obtained using CHAMP and GRACE POE data. The trends in how POE density estimates varied for all four satellites were found to be the same or similar. The comparisons were made for different baseline atmospheric density models, different density and ballistic coefficient correlated half-lives, and for varying levels of solar and geomagnetic activity. The comparisons in this research help in understanding the variation of density estimates for various satellites with different altitudes and orbits.