中文摘要：

电离层掩星数据反演的传统方法是Abel反演法及其改进方法，而实际电离层的非球对称给电离层电子密度反演带来误差．本文研究了在气象领域广泛应用的变分同化方法在掩星数据反演电离层电子密度廓线的可行性，利用IRI和Nequick模型模拟掩星真值场和背景场进行数值模拟反演，并与Abel反演法反演结果进行对比．结果表明，变分同化反演方法能够有效综合模式和观测数据，使得反演结果精度较高；与Abel反演法相比，反演的电子密度廓线F2层峰值浓度误差在10％以下，而Abel反演法在20％～30％之间；而且变分N化反演法对误差扰动有较好的过滤性，因此实用性较强．

英文摘要：

GPS radio occultation measurements are brand new and highly effective , which can detect atmosphere and ionosphere, but still in the development and perfection. The Abel inversion method and the improved method is usually applied in the inversion of ionosphere occultation data , which could cause error because of the spherical asymmetry which is not in the real ionosphere. In this paper the variational assimilation method which is widely used in the field of meteorology is introduced for the inversion of electron density profile from the radio occultation data. A numerical simulation is carried out to validate the variational assimilation method. The simulated data include background field and observation field which are simulated by the IRI model and Nequick model. The inversion results are compared with the results inversed by the Abel inversion method. The results show that the variational assimilation inversion method can integrate the model results and observations and the inversion results are quite similar with the true values simulated by the IRI. Compared with the Abel inversion method, the corelative error of the retrieved peak value of the F2 layer below 10%, while the Abel inversion method between 20%- 30%. The variational assimilation inversion method can obtain a better result when there is a error disturbance , this make the method more practical. But the background and observation error matrix still need a improvement to get a better results.