极地研究

极地研究 ›› 2025, Vol. 37 ›› Issue (4): 775-785.DOI: 10.13679/j.jdyj.20240024

所属学科：极区空间物理

• 研究论文 • 上一篇    下一篇

基于函数型主成分分析的极区电离层ELDI特征聚类研究

武业文1,2,3, 金婷婷1, 刘瑞源4, 陈鹏蕾1, 贡康1, 陈涵怡1   

  1. 1南京信息工程大学数学与统计学院, 江苏 南京 210044;
    2南京信息工程大学江苏省应用数学中心, 江苏 南京 210044;
    3南京信息工程大学江苏省系统建模与数据分析国际合作联合实验室, 江苏 南京 210044;
    4自然资源部极地科学重点实验室, 中国极地研究中心(中国极地研究所), 上海 200136
  • 收稿日期:2024-03-11 修回日期:2024-05-14 出版日期:2025-12-30 发布日期:2026-01-12
  • 通讯作者: 贡康
  • 基金资助:
    国家自然科学基金项目和中国航天科技集团有限公司第八研究院产学研合作基金资助项目资助

Clustering of E-Layer dominated ionosphere (ELDI) characteristics in the polar regions based on functional principal component analysis

WU Yewen1,2,3, JIN Tingting1, LIU Ruiyuan4, CHEN Penglei1, GONG Kang1, CHEN Hanyi1   

  1. 1School of Mathematics and Statistics, Nanjing University of Information Science and Technology, Nanjing 210044, China;
    2Center for Applied Mathematics of Jiangsu Province, Nanjing University of Information Science and Technology, Nanjing 210044, China;
    3Jiangsu International Joint Laboratory on System Modeling and Data Analysis, Nanjing University of Information Science and Technology, Nanjing 210044, China;
    4Key Laboratory of Polar Science, Ministry of Natural Resources, Polar Research Institute of China, Shanghai 200136, China
  • Received:2024-03-11 Revised:2024-05-14 Online:2025-12-30 Published:2026-01-12

PDF (PC)

104

摘要： 针对极区电离层特殊的E层占优电离层(ELDI)电子密度分布特征, 本文基于2007—2010年南半球冬季COSMIC卫星在南极地区的观测数据, 采用函数型主成分分析法提取出ELDI事件的两个主成分, 并应用k-means聚类方法实现了ELDI特征的分类识别。研究结果表明, 不同类别的ELDI在电子密度剖面形状上存在明显单峰和双峰差异性, 其高发区随地磁纬度和磁地方时的分布也存在明显不同。进一步对比分析显示, 不同类别ELDI的电离层特征参量具有统计学显著性差异(p<0.05)。

Abstract:

The E-Layer Dominated Ionosphere (ELDI) is characterized by high electron density in the E-layer and occurs primarily in the polar regions. This paper used data from the Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) satellite to analyze the ionospheric electron density in the Antarctic region during winter for 2007—2010. Functional principal component analysis was used to identify the two principal components of ELDI events. k-means clustering was carried out to categorize the events. The results show that categories differ in terms of electron density profiles: some are bimodal while others are unimodal. Categories also differ significantly in terms of the geomagnetic latitude and time of dominant events. The distributions of the characteristic ionospheric parameters of different types of ELDI events were compared and analyzed and clear differences were found (p<0.05).


Key words: functional principal component analysis, cluster analysis, E-layer dominated ionosphere,  electron density profile, Antarctic