Development of drag-coefficient models for more accurate orbit prediction and thermospheric density estimation

Abstract:

An accurate modeling of the atmospheric drag force parameters - density and drag coefficient, is a key component of both orbit prediction and scientific investigations into the Earth's atmosphere. With the object population in certain low Earth orbit (LEO) belts reaching a critical spatial density, accurate real-time monitoring as well as prediction of satellite positions into the future has emerged as a problem of utmost importance for orbit maintenance and collision avoidance. Uncertainties in atmospheric drag remain the foremost contributor of prediction errors in the low altitude LEO regime with the drag-coefficient playing a dominant role. Additionally, assumptions pertaining to the drag coefficient are detrimental to atmospheric densities derived from satellite decay data. Biases introduced into satellite-derived thermospheric densities feed into semi-empirical atmospheric models used for orbit determination as well as long-term density trends that provide insights into the Earth’s atmosphere. In most orbit determination applications, a cannonball model is used such that drag-coefficient variations are averaged out. On the other hand, high-fidelity physics-based drag-coefficient models or gas-surface interaction models (GSIMs) are capable of capturing drag-coefficient variations but are inhibited by a lack of accurate knowledge of their input parameters. We report on the development of a new drag-coefficient modeling approach based on spatial and temporal Fourier series expansions that can be used to improve orbit determination and prediction, estimates of atmospheric densities as well as input parameter space of physics-based drag-coefficient models. Such an approach allows the estimation of time-variations in the drag coefficient specific to changes in attitude of the satellite and/or periodic changes in governing ambient parameters. Since the Fourier coefficients, being orthogonal to each other, can be estimated from tracking data, accurate knowledge of gas-surface interaction parameters is not required. In the fall meeting, we will outline the theoretical framework of these models and discuss their applications from orbit determination to density inversion from satellite data.