Scholarlink

<p>Yunnan and Sichuan provinces in China are situated within the northern crest of the Equatorial Ionization Anomaly (EIA), a region where ionospheric disturbances during geomagnetic storms are particularly pronounced. To address this, the study utilizes ionospheric data collected during geomagnetic storms from 2011 to 2018 at 48 stations of the Crustal Movement Observation Network of China (CMONOC). By employing the Sparrow Search Algorithm (SSA) in conjunction with a ConvLSTM-BiLSTM neural network leveraging multi-channel features, a localized low-latitude ionospheric storm model was developed. During the main and recovery phases of the 2017 and 2018 geomagnetic storms, the model achieved root mean square errors (RMSE) of 1.33 TECu and 1.44 TECu for 2017, and 1.22 TECu and 1.16 TECu for 2018, respectively. Significantly outperforming existing models, such as LSTM-CNN, ConvLSTM, and the Global Ionosphere Map (GIM), the model demonstrates superior performance. Results demonstrate that the SSA-ConvLSTM-BiLSTM model effectively and reliably predicts the spatiotemporal evolution of the ionosphere over the Yunnan-Sichuan region during severe geomagnetic storms.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Storm-time ionospheric model over Yunnan-Sichuan area of China based on the SSA-ConvLSTM-BiLSTM algorithm

  • 期刊论文

  • 发表日期：

    2025年3月10日

查看全文
  • Wang Li,
  • Haoze Zhu,
  • Fangsong Yang,
  • Can Wen,
  • Shuangshuang Shi,
  • Dongsheng Zhao,
  • Changyong He,
  • Zhen Li

摘要

Yunnan and Sichuan provinces in China are situated within the northern crest of the Equatorial Ionization Anomaly (EIA), a region where ionospheric disturbances during geomagnetic storms are particularly pronounced. To address this, the study utilizes ionospheric data collected during geomagnetic storms from 2011 to 2018 at 48 stations of the Crustal Movement Observation Network of China (CMONOC). By employing the Sparrow Search Algorithm (SSA) in conjunction with a ConvLSTM-BiLSTM neural network leveraging multi-channel features, a localized low-latitude ionospheric storm model was developed. During the main and recovery phases of the 2017 and 2018 geomagnetic storms, the model achieved root mean square errors (RMSE) of 1.33 TECu and 1.44 TECu for 2017, and 1.22 TECu and 1.16 TECu for 2018, respectively. Significantly outperforming existing models, such as LSTM-CNN, ConvLSTM, and the Global Ionosphere Map (GIM), the model demonstrates superior performance. Results demonstrate that the SSA-ConvLSTM-BiLSTM model effectively and reliably predicts the spatiotemporal evolution of the ionosphere over the Yunnan-Sichuan region during severe geomagnetic storms.