基于银川电离层垂测仪电子浓度反演的一次强电离层暴观测

Based on Electron Density Inversion of Yinchuan Vertical Ionosonde

根据银川电离层垂测仪的回波数据,采用脉冲压缩技术,使用Bernoulli映射序列对发射信号进行编码,解决实际探测中回波信号混合强杂波干扰的问题,从而获得高质量频高图.为从图中提取电离层的关键信息,将信号处理问题转换成计算机视觉中的语义分割任务,构建原始频高图数据集,并进行离散化和人工标注等预处理.通过训练c GAN神经网络分析得到频高图中各层回波的特征参数,达到分割不同描迹的目的.采用改进式国际参考电离层底部反演模型和NeQuick顶部模型对垂测仪上空的电子浓度剖面进行反演,根据张衡一号卫星的实测数据对顶部的计算结果进行修正.通过将计算得到的总电子浓度与CDDIS公开的数据结果对比,验证了垂测仪数据的准确性.在此基础上,结合高沙窝磁通门的地磁数据,垂测仪于2023年4月23-24日的大地磁暴期间成功观测到电离层异常变化的全过程并给出了总电子浓度变化结果,为探究中国西部电磁环境变化提供准确可靠的观测数据.

This study is based on the echo data from the Yinchuan vertical ionosonde.The ionosonde supports scanning in the frequency range from 1 to 30 MHz,with a distance resolution of 1.5 km and a reception window ranging from 67.5 km to 560.1 km.It utilizes pulse compression technology and encodes the transmission signal using Bernoulli mapping sequences,successfully resolving the issue of echo signal mixture with strong clutter interference in practical detection,thus obtaining Ionograms of high quality.In order to extract key information of ionosphere from the ionograms,the signal processing problem is transformed into a semantic segmentation task in computer vision,constructing an original ionograms dataset,and undergoing preprocessing such as discretization and manual annotation.By training a cGAN neural network to analyze the characteristic parameters of each layer’s traces in the ionograms,the goal of segmenting different traces is achieved.The network is suitable for processing various types of ionograms under calm conditions,with an accuracy rate of over 95%,effectively saving time in manual parameter measurement and improving processing efficiency.An improved bottom inversion model of the International Reference Ionosphere and the NeQuick top model is used to invert the electron density profile above the ionosonde,while the top calculation results are corrected according to the actual measurement data from“CSES-1”.By comparing the total electron content calculated with the data results publicly available from CDDIS,the accuracy of the ionosonde data is verified.On this basis,combined with the geomagnetic data acquired by the Gaoshaowo magnetometer,the ionosonde successfully observed the entire process of ionospheric anomalies during the geomagnetic storm on 23–24 April,2023,and provided the results of the total electron content changes,offering accurate and reliable observational data for exploring the electromagnetic environment changes in western China.