基于电子密度同化的热层大气密度预报方法

采用热层电离层耦合模式TIEGCM和集合卡尔曼滤波同化方法，利用同化COSMIC电离层掩星电子密度数据优化热层电离层参量，并将模式预报的大气密度与CHAMP卫星大气密度数据进行对比，分别开展模拟和实测数据的同化预报实验.在模拟数据同化实验中，状态向量包含温度、风场和离子成分的实验结果表明，仅优化温度即可达到最优的热层大气密度预报效果.在实测数据同化实验中，将温度作为状态向量参数，优化结果表明，循环同化过程中模式预报的大气密度相对偏差的均方根误差在48h内从38%减小到27%，同化稳定时间至少需要30h.预报过程中大气密度预报效果的改善持续时间为34h.这表明电子密度同化能够改善热层大气密度的预报精度，设计的实验方案合理可行，可获得较长的预报时效. 

Thermospheric Density Prediction Based on Electron Density Assimilation

Using the thermosphere ionosphere coupling model TIEGCM and the thermospheric and ionospheric observations, the assimilation and forecast experiments with simulated and measured data are carried out based on the ensemble Kalman filter method respectively. The results of the simulated assimilation experiments with different thermospheric ionospheric parameters show that the temperature is the key parameter to improve the thermospheric density. In the assimilation experiments, the temperature is taken as the parameter of state vector. The optimization results show that the root mean square error of relative deviation of atmospheric density predicted by the model is reduced from 38% to 27% in 48 hours, and the stabilization time of assimilation is at least 30 hours. However, it may need at least 30 hours to achieve the best assimilation results when only the temperature is estimated, and the e-folding time of neutral density is 34 hours. The accuracy of the neutral density prediction in the TIEGCM has been significantly improved, which indicates that it is feasible to improve the prediction accuracy of the neutral density of the thermosphere using ionospheric observations