DOPE: Dynamic Orbit Propagation for RealisticUncertainty Characterization in Low Earth Orbit

Covariance realism in low Earth orbit (LEO) operations is critical for ensuring space safety. Realism requires both precise and accurate characterization of the vehicle dynamics and environment. This study focuses on modeling the temporal auto-correlation (half-life) in atmospheric density and associated uncertainty (both due to density model and space weather drivers), in real-time operations. Atmospheric density being a major perturbing force in LEO significantly impacts precise orbit prediction and uncertainty quantification. Current orbit determination (OD) / orbit propagation (OP) tools neglect the temporal auto-correlation component in atmospheric density and assume density to be deterministic with associated uncertainty modeled as process noise. This research presents a novel technique, named the Dynamic Orbit Propagator and Estimator (DOPE), to incorporate uncertainty in the atmospheric density model within state propagation and capture the temporal auto-correlation dynamics as a first-order Gauss-Markov (FOGM) process. This enables a realistic representation of density half-life within the real-time OD/OP framework. In DOPE, the state vector incorporates density perturbations ($\delta\rho$), preserving the standard propagation architecture while maintaining computational efficiency and operational practicality. This work forms part of the next-generation atmospheric drag modeling framework under development through efforts supported by the Intelligence Advanced Research Projects Activity (IARPA) Space Debris Identification and Tracking (SINTRA) program and the Office of Space Commerce (OSC). The technique presented here is a critical component of this framework. Monte Carlo (MC) simulations validate that varying half-life values produce statistically consistent orbital deviations, enhancing realism in orbit prediction, conjunction assessment, and reliable tracking of Resident Space Objects (RSOs).