Ryan Blay, Orion Space Solutions; Connor Johnstone, Orion Space Solutions; Jeremy Highley, Orion Space Solutions; Junk Wilson, Orion Space Solutions; John Noto, Orion Space Solutions; Geoff Crowley, ASTRA LLC.
Keywords: Satellite drag, Thermospheric density, Data assimilation, Space Weather, Space Environment, Nowcast, Forecast, Ensemble Kalman Filter
Abstract:
A significant challenge of global Space Domain Awareness is the sheer number of objects that are now in orbit around Earth. As the demand for satellites increases, and the density of satellites in space increases, there is a need for more robust space domain awareness capabilities, specifically the prediction of satellite orbits. Perturbing forces which change the orbital trajectory of space objects can make this process difficult, resulting in greater uncertainty around collision avoidance and maneuvering, causing optical sensors to spend additional time searching for objects they are tasked to track. When these forces are modeled or represented incorrectly, errors in orbital position accumulate, potentially causing ambiguity between objects for optical sensors. Atmospheric drag is especially difficult to specify for the Low Earth Orbit (LEO) regime (below 2000 km altitude), yet this is where most satellites reside.
The accuracy of current modeling of space object density is unsophisticated enough that it does not provide estimates of the uncertainty regarding drag or object position. More accurate modeling can increase accuracy, decrease search volumes, and thus increase the capacity of individual sensors to make additional observations. For satellite operators, conjunction assessment and collision avoidance are daily issues which depend upon having high accuracy orbit estimates with well-characterized uncertainties and recent observations. Any loss in orbit estimate accuracy significantly increases the chances of a catastrophic collision. Existing atmospheric drag models are not robust enough to provide accurate predictions and cannot take advantage of additional available data, or improved forecasting techniques.
To truly achieve actionable and timely collision avoidance processes, improved atmospheric density and drag models are required. Orion Space Solutions has developed a drag specification and forecast tool called Dragster based on well-validated, full-physics atmospheric models and ensemble data assimilation techniques.
The Dragster ensemble model is based on several empirical models as well as three well-validated Global Circulation Models: (a) the Thermosphere Ionosphere Electrodynamics Global Circulation Model (TIE-GCM), (b) the Thermosphere Ionosphere Mesosphere Electrodynamics Global Circulation Model (TIME-GCM) which includes coupling into the mesosphere, and (c) the Coupled Thermosphere Ionosphere Plasmasphere electrodynamics model. The neutral atmosphere codes solve the non-linear momentum, energy, and composition equations time-dependently over the globe, to provide neutral dynamics, temperature, and the distribution of neutral species. The three-dimensional distribution of neutral density is obtained from the temperature and composition, which together with the neutral winds provide the necessary parameters for satellite drag prediction. The self-consistent ionosphere is also important and necessary to ensure accurate conductivities needed for characterizing high latitude Joule heating and momentum forcing as well as for ion drag and realistic wind determination.
Dragster is an ensemble assimilative framework incorporating both first principles (full-physics) and empirical models. Each model type is used to perform ensemble assimilation and hence the various models will sometimes be referred to as super-ensemble members. The software design is modular, such that as improved atmospheric models evolve, they can be substituted into Dragster for evaluation/testbed purposes. This forward-looking design means that the Dragster tool can be optimized to leverage an ensemble of models most appropriate for the task at hand, and future improvements to any such models will flow directly into Dragster performance. For example, several whole atmosphere models are currently being developed that could be introduced into the Dragster framework. Dragster propagates the model ensemble members forward to predict the most probable thermospheric state and its uncertainty based on inter-model differences, similar to the familiar plots of hurricane landfall with uncertainties. It must be kept in mind that unlike tropospheric weather, the thermosphere is strongly driven by external inputs and depends less on the current and prior states. The external drivers are the solar and geomagnetic forcing.
Goals of this research include provision of density/drag uncertainty calculations, validation metrics, additional validation testing, feasibility of assimilating new datasets and interfacing to standard astrodynamics packages. These capabilities are especially attractive as any technology in this space must be able to interface with and/or assimilate data from a variety of stake holders. Dragster can be programmed to interface with or assimilate data from a variety of key stakeholders including civil and government agencies (NASA), military (Space Force, Air Force), and commercial entities (SpaceX, Amazon).
The expected outcome of the Dragster development effort is to provide a validated ensemble density model approach that out-performs the current state of the art.
Date of Conference: September 19-22, 2023
Track: Atmospherics/Space Weather