Thermospheric basis functions for improved dynamic calibration of semi-empirical models

State-of-the-art satellite drag models require upgrades to meet operational Precision Orbit Determination requirements for collision avoidance, reentry predictions and catalog maintenance. However, accurate model representations of thermospheric density are not currently possible without the use of data assimilation, or model calibration. Furthermore, due to sparse data sampling in the thermosphere, such calibration has only been successfully demonstrated by fitting the available data to a low-dimensional model. The High Accuracy Satellite Drag Model (HASDM), used operationally by the Space Surveillance Network to aid in the tracking of low-earth orbiting satellites, compensates for errors in the Jacchia-70 static diffusion model by fitting a truncated set of spherical harmonics to a subset of recent satellite tracking data. We present a technique to derive a set of basis functions better suited to capturing the spatial variability and response of the thermosphere. By comparing the uncompensated Jacchia-70 model with the Thermosphere-Ionosphere-Electrodynamic General Circulation Model (TIEGCM), we create a new set of orthogonal basis functions that can be used to calibrate semi-empirical models such as HASDM with increased accuracy in the presence of sparse data. An initial analysis of the new approach, driven by synthetic data, shows a 32.9% improvement in the RMS error over the current implementation of HASDM.