一种全球临近空间大气密度建模方法及应用

基于TIMED/SABER卫星2002—2018年观测的20~100 km大气密度数据，统计获得多年月平均值和标准偏差的全球网格数据。利用网格数据，分析了大气密度的变化特征。以网格数据为基准，计算了USSA76的相对偏差，分析了USSA76相对偏差的分布特征。以网格数据为驱动，将大气密度表征为平均值与大尺度扰动量和小尺度扰动量的加和，大尺度扰动和小尺度扰动分别采用余弦函数和一阶自回归模型表征，初步建立了全球临近空间大气密度模型。通过对比模型仿真值与激光雷达观测值，表明模型仿真值与观测值具有较好的吻合度，验证了建模方法的可行性。利用蒙特卡罗方法可再现给定轨迹上所有可能的大气状态。 

A modeling method and its application of global atmospheric density in near space

Based on the 20-100 km atmospheric density data observed by TIMED/SABER satellite from year 2002 to 2018, the global grid data of monthly average and standard deviation are calculated statistically. Based on the grid data, the characteristics of atmospheric density variation are analyzed. The relative errors of USSA76 are calculated and the distribution characteristics of relative errors of USSA76 are analyzed. In addition, driven by the grid data, the atmospheric density is characterized as the sum of the monthly average and the large-scale disturbances and small-scale disturbances. The large-scale perturbations and small-scale perturbations are characterized by cosine functions and first-order autoregressive models, respectively. And a global near-space atmospheric density model is initially established. By comparing the simulated values of the model with the observed values of the lidar, the results show that the model values have a good agreement with the observed values, which verifies that the modeling method is feasible. Finally, Monte Carlo method can be used to reproduce all possible states of atmospheric density on a given trajectory.