The space weather research community relies heavily on thermospheric density data to understand long-term thermospheric variability, construct assimilative, empirical, and semi-empirical global atmospheric models, and validate model performance. One of the challenges in resolving accurate thermospheric density datasets from satellite orbital drag measurements is modeling appropriate physical aerodynamic drag force coefficients. The drag coefficient may change throughout the thermospheric environment due to model dependencies on composition and altitude. As such, existing drag coefficient model errors may be altitude and solar cycle dependent, with greater errors at higher altitudes around 500 km near the oxygen-to-helium transition region. This can lead to errors in orbit-derived density datasets and models. In this paper, inter-satellite density comparisons at ~500 km are evaluated to constrain drag coefficient modeling assumptions. Density consistency results indicate that drag coefficient models with incomplete energy and momentum accommodation produce the most consistent densities, while the standard diffuse modeling approach may not be appropriate at these altitudes. Models with momentum accommodation between 0.5 - 0.9 and energy accommodation between 0.83 - 0.96 may be the most appropriate at upper thermospheric altitudes. Modeling drag coefficients with diffuse gas-surface interactions could lead to errors in derived density of ~25% and in-track satellite orbit prediction uncertainty during solar maximum conditions on the order of hundreds of meters.

Atmospheric drag describes the main perturbing force of the atmosphere on the orbital trajectories of near-Earth orbiting satellites. The ability to accurately model atmospheric drag is critical for precise satellite orbit determination and collision avoidance. Assuming we know atmospheric winds and satellite velocity, area and mass, the primary sources of uncertainty in atmospheric drag include mass density of the space environment and the spacecraft drag coefficient, CD. Historically, much of the focus has been on physically or empirically estimating mass density, while CD is treated as a fitting parameter or fixed value. Presently, CD can be physically modeled through energy and momentum exchange processes between the atmospheric gas particles and the satellite surface. However, physical CD models rely on assumptions regarding the scattering and adsorption of atmospheric particles, and these responses are driven by atmospheric composition and temperature. Modifications to these assumptions can cause CD to change by up to ~40%. The nature and magnitude of these changes also depend on the shape of the spacecraft. We can check the consistency of the CD model assumptions by comparing densities derived from satellite drag measurements and computed CD values for satellites of different shapes orbiting in the same space environment. Since all of the satellites should see the same density, offsets in the derived densities should be attributable to inconsistencies in the CD model. Adjusting the CD model scattering assumptions can improve derived density consistency among the different satellites and inform the physics behind CD modeling. In turn, these efforts will help to reduce uncertainty in CD, leading to improved atmospheric drag estimates.