摘要
针对通过GNSS数据进行电离层层析时真实数据不足会导致层析结果的分辨率越高,模型精度反而越低的问题,提出一种混合格网层析方法。该方法是在代数重构法层析的基础上,将电离层划分为多层,每层采用不同的分辨率做层析,之后对调各部分的分辨率再层析,将相同分辨率下的格网电子密度组合,作为最终结果。实验发现,与单一格网相比混合格网层析格网电子密度精度明显提高,3种方法的均方根误差都降低了10%以上,其中ART方法降低了20.3%。对比使用同一迭代方法在不同的格网划分条件下的结果,发现3层划分和4层划分普遍比2层划分精度高,ART的精度提高了30%以上。
When ionospheric tomography was performed with GNSS data,the real valve of grid ionospheric electron density was lacked.It caused that the higher the resolution,the lower the accuracy.In response to this problem,a hybrid grid tomography method is proposed.This method was based on algebraic reconstruction tomography,which divided the ionosphere into multiple layers;each layer used different resolution tomography,and then reverses,Combining the electron density of grid with the same resolution as the final result.The experiment find that compared with a single grid,the accuracy of the electron density of the hybrid grid tomographic grid is significantly improved,and the root mean square error of the three methods was reduced by more than 10%,of which the ART method is reduced by 20.3%.Comparing the results of using the same iterative method under different grid partitioning conditions,it is found that the accuracy of three-level and four-level partitions is generally higher than that of two-level partitions,and the accuracy of ART is improved by more than 30%.This paper proposes a scheme based on hybrid grid tomography,which effectively improves the accuracy of grid electron density,and explores the relationship between grid division and tomographic results.
作者
孙嗣文
余接情
SUN Siwen;YU Jieqing(Wuhan University,Wuhan 430070,China;China University of Mining and Technology,Xuzhou,Jiangsu 221000,China)
出处
《测绘科学》
CSCD
北大核心
2020年第7期33-37,共5页
Science of Surveying and Mapping