利用卫星两行轨道根数反演热层密度

两行轨道根数（TLEs）是基于一般摄动理论产生的用于预报地球轨道飞行器位置和速度的一组轨道参数，通过求解大气阻力微分方程，可反演出热层大气密度. 本文选取近圆轨道CHAMP卫星和椭圆轨道Explorer8卫星，以两行轨道根数数据为基础，计算反弹道系数，并根据不同轨道特征采用两种不同反演方法对热层大气密度进行研究. 结果表明，这两种方法反演得到的大气密度与实测值均符合较好，其中CHAMP卫星的反演结果和经验模式值相对于实测值的误差分别为7.94%和13.94%，Explorer8卫星的误差分别为9.04%和14.32%. 相比模式值，利用两行轨道根数数据反演的热层大气密度更接近于实测值，说明该方法可以作为获取大量可靠大气密度数据的一种有效途径. 

Research on Thermospheric Densities Derived from Two-line Element Sets

Two-line Orbital Element Sets (TLEs) consist of mean orbital elements at epoch, along with the NORAD (North American Aerospace Defence Command) catalog number, international designator, epoch and additional fitting parameters. These information can be used to derive thermospheric densities through integration of differential equation for mean motion. For near-circular orbit satellites, derived thermospheric density can be seen as real density because of their stable orbit height, while for elliptical orbit satellites, thermospheric density at perigee and apogee can be different as much as several orders. So different methods were applied to derive thermospheric density according to different satellite orbits. This paper chooses CHAMP and Explorer 8 satellites, whose orbits are respectively near-circular and elliptical, as our research cases. The inverse ballistic coefficient B (B-factor) was firstly derived based on TLEs data, then thermospheric densities were derived with different methods according to different orbit characters. Finally, a comparison was made among TLEs-derived density, NRLMSISE-00 model density and observed (or reference) density. The result shows that the average error of TLEs-derived density and empirical model density with respect to observed value for CHAMP is 7.94% and 13.94% respectively, and the average error with respect to reference value for Explorer 8 is 9.04% and 14.32% respectively. This result indicates that TLEs-derived density is closer to the real density than empirical model density, and this method provides an effective way to obtain extensive and reliable atmosphere density data.