Calibrating the scale of the NRLMSISE00 model during solar maximum using the two line elements dataset

Empirical mass density models for the thermosphere are widely used for object orbit determination and prediction, object collision avoidance, and re-entry analysis. But the error of these empirical models can often reach 15-30% or even larger during highly disturbed periods of the space environment. On the other hand, the mass density of the thermosphere can be derived from the orbit information contained in the two line elements (TLE) dataset. This technique provides an approach for calibrating the empirical model. Here we select TLE data of 36 low Earth orbiters (LEOs) recorded during 2000-2002 (solar maximum). The ratios of the TLE-derived densities to those from the empirical NRLMSISE00 model are calculated and used to calibrate the scale error of the NRLMSISE00 model by applying a linear height-dependent function. The calibration models for the NRLMSISE00 model during 2000-2002 are then obtained by a least squares adjustment procedure. The calibration factors at 250, 400, and 550 km from this calibration model are compared to the density ratios obtained by Emmert et al. (2008) who used TLE data from ∼5000 LEOs. The result indicates that the biases between these two independent factors at the 3 altitudes are all within +/-2%, and the standard deviations (STDs) are under 7%. Another 5 LEOs with altitudes ranging from 200 to 500 km are also selected to validate the precision of the calibration model. Their density ratios are calculated using the calibration model and the NRLMSISE00 model, respectively. The results demonstrate that by applying this calibration method the relative root mean square (RMS) error of the NRLMSISE00 model can be reduced by about 9%.