Precise thermospheric mass density modelling for orbit prediction of low earth orbiters

He, C 2019, Precise thermospheric mass density modelling for orbit prediction of low earth orbiters, Doctor of Philosophy (PhD), Science, RMIT University.

Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
He.pdf Thesis application/pdf 29.75MB
Title Precise thermospheric mass density modelling for orbit prediction of low earth orbiters
Author(s) He, C
Year 2019
Abstract The steady increase in the number of space objects near the Earth has raised critical security concerns for the low Earth orbit (LEO) space environment where most of the near-Earth satellites missions operate. Orbit prediction (OP) is the foundation of many space missions and applications in LEO, e.g., space situational awareness, re-entry prediction and debris removal. However, the precision of OP is limited due to the accuracy of thermospheric mass density (TMD) prediction. In the past few decades, more atmospheric data sets have been inferred from different techniques such as the Global Navigation Satellite System, satellite laser ranging and two-line-element catalogue. However, accurately predicting TMD is still a challenging task due to the limited knowledge of thermospheric dynamics and the lack of measurements with sufficient temporal and spatial resolution.

In this research, a precise OP platform for the analysis and prediction of the orbital motion of satellite and and space debris is developed. It consists of various precise perturbation models of gravitational and non-gravitational forces. This includes the high-order Earth gravitational acceleration with the effect of solid and ocean tides, third-body perturbations from other celestial bodies in the solar system, the general relativity effects, aerodynamic acceleration, direct solar radiation pressure, and Earth's albedo and infrared radiation pressure. Coordinate transformation is established on the precise time systems and the measured Earth orientation parameters. The developed OP platform is validated against the historical precise orbits of LEO satellites.

In order to evaluate the most representative classes of empirical TMD models, a comprehensive comparison of 12 models is performed. The vertical variability, horizontal scale and the capability to capture the physics-based features of the selected models are investigated. Various validations against the TMD estimated from on-board accelerometer measurements of the GRACE satellites have been conducted. The performance of these models in the OP of the GRACE-A satellite is assessed under different solar and geomagnetic conditions. Also discussed is the coupling effect between the TMD and ballistic coefficient that measures the impact of atmospheric friction on the space object.

The impact of TMD variations on orbit dynamics of LEO objects is an important focus in this thesis, which has not been well-quantified in previous studies. Intra-annual, intra-diurnal and horizontal TMD variations are reproduced using the empirical model DTM-2013. Also evaluated are physics-based variations including the equatorial mass density anomaly (EMA) and midnight mass density maximum (MDM), which exhibit both temporal and spatial variations and are simulated by the Thermosphere Ionosphere Electrodynamics General Circulation Model. The analysis is based on the one-day OP simulation at 400 km. The result show that TMD variations have a dominant impact on the predicted orbits in the along-track direction. Semiannual and semidiurnal TMD variations exert the most significant impact on OP among the intra-annual and intra-diurnal variations, respectively. In addition, both EMA and MDM create weaker but still discernible impacts than other TMD variations. Some recommendations for TMD modelling are also presented.

Moreover, precise modelling of TMD during geomagnetic quiet time is performed. This is undertaken using the TMD data inferred from GRACE (500 km), CHAMP (400 km) and GOCE (250 km) satellites during the year of 2002-2013. Three different methods including the Fourier analysis, spherical harmonic (SH) analysis and the artificial neural network (ANN) technique are adopted and compared in order to determine the most suitable methodology for the TMD modelling. Additionally, different combinations of time and coordinate representations are also examined in the TMD modelling. The results reveal that the precision of the low-order Fourier-based model can be improved by up to 10% using the geocentric solar magnetic coordinate. Both the Fourier- and SH-based models have drawbacks in approximating the vertical gradient of TMD. The ANN-based model, however, has the capability in capturing the vertical TMD variability and is not sensitive to the input of time and coordinate.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Science
Subjects Geodesy
Navigation and Position Fixing
Keyword(s) Thermospheric Mass Density
Empirical Thermospheric Mass Density Model
Low Earth Orbit
Orbit Prediction
Artificial Neural Network
Version Filter Type
Access Statistics: 85 Abstract Views, 131 File Downloads  -  Detailed Statistics
Created: Thu, 29 Aug 2019, 11:46:53 EST by Adam Rivett
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us