Machine Learned Atmospheric Force Model Trained with Two Line Elements

For objects orbiting in Low Earth Orbit (and in the absence of maneuvers), the atmospheric drag force is the dominant perturbation force that affects the trajectory over time. Thus, accurately modeling a time and state dependent atmosphere has been a high priority to enable high accuracy propagations and predictions of orbiting objects in Low Earth Orbit.

Over each time span and across multiple missions, modeling errors can be organized to be the training data to a machine learning process that refines the atmospheric model stochastically to reduce prediction errors.