Neutral densities from Two-Line-Elements for operational forecasting of the thermosphere

Forecasting neutral densities in the thermosphere is important for quantifying the atmospheric drag a satellite experiences. This was demonstrated in the February 2022 “Starlink storm”, which saw SpaceX lose 38 of 49 recently launched satellites after under-estimating the drag caused by a Kp 5+ storm. There are no sources of direct, near real-time thermospheric density observation. Instead, neutral densities are derived from other observations, with the most common methodology being via satellite tracking data in the form of Two-Line Elements (TLEs). TLEs have previously been used for obtaining neutral densities for historic densities, rather than near-real time.
We present the methodology used to operationally obtain near-real time neutral densities from TLEs, which can then be assimilated by AENeAS (Advanced Ensemble Ne Assimilation System) to allow forecasting of the density in Low Earth Orbit (LEO). As TLEs for an object become available, they are checked for outliers and manoeuvres by the Tracking Real-time Unusual Satellite Trajectories (TRUST) algorithm. This results in pairs of TLEs with a drop in semi-major axis which can be related to the average density experienced. The drop in semi-major axis is also related to the ballistic coefficient of the object (its shape and area-to-mass ratio). This is accounted for by looking at “reference objects” with known ballistic coefficients, providing the best possible neutral densities. Information on the known ballistic coefficients can then be passed to less well-behaved objects in similar orbits, albeit with increasing errors as the orbits become less similar and objects less spherical in shape.