Using differential satellite drag as an experiment to characterise gas–surface interactions

Abstract Uncertainty in atmospheric density models and drag coefficient modelling contributes to orbit prediction errors for satellites in Low Earth Orbit (LEO). It is of interest to better characterise the Gas–Surface Interactions (GSI) to improve drag coefficient modelling, which is, however, hindered by a lack of dedicated in-orbit experiments. We propose a new experiment to estimate the energy accommodation coefficient of the Diffuse Reflection with Incomplete Accommodation (DRIA) GSI model. The experiment consists of two small satellites with Global Navigation Satellite Systems (GNSS) receivers and attitude determination systems to derive atmospheric density observations from the positioning data. The experiment has two key features. The first is the satellites’ close along-track formation flying, such that they should observe the same atmospheric density with a slight delay due to their along-track separation. Second, the satellites have controllable panels to modify their drag coefficients’ response to GSI substantially. Hence, the satellites’ atmospheric density observations will agree only when the DRIA model’s energy accommodation coefficient is selected correctly. We demonstrate by simulation that the energy accommodation coefficient can be estimated at least once daily with a precision of 5–10% for satellites with decimeter-accuracy GNSS positioning. Given that GNSS receivers and attitude determination systems are common for small satellites currently in LEO, we conclude that there are plenty of opportunities to utilise existing data for the proposed experiment. Valuable byproducts would be atmospheric density observations that are relatively free of systematic errors.