Scholarlink

<p>The search for reliable earthquake precursors remains one of the most challenging problems in geophysics. The lithosphere–atmosphere–ionosphere coupling (LAIC) mechanism has been widely proposed to explain reported pre-seismic anomalies in total electron content (TEC), Outgoing Long Wave Radiation (OLR), and land surface temperature (LST); however, the consistency and generalizability of these anomalies across various scenarios remain unclear. Here, we introduce a novel multi-parameter Seismo-Atmospheric Anomaly Index (SAI) that systematically integrates GNSS-based TEC, Swarm satellite TEC, MODIS-derived Land Surface Temperature (LST), and Outgoing Long Wave Radiation (OLR) into a single normalized score. A Nonlinear Auto Regressive with eXogenous inputs (NARX) machine learning model is employed to detect anomalies in OLR and LST. By incorporating a sliding 5-day time window, the SAI framework is explicitly designed to capture lagged or asynchronous anomalies across different atmospheric and ionospheric layers. Applying this method to five moderate-to-large Indonesian earthquakes (Mw 6.6–7.1, 2020–2023), we show that only the Mw 6.9 Bengkulu earthquake (18 August 2020) exhibited clear, multi-parameter precursors exceeding the anomaly detection threshold, while no significant anomalies were detected for the remaining four events. This dual outcome detecting positive anomalies in one case while ruling them out in others provides a rare quantitative evaluation of LAIC and highlights that these seismo-atmospheric precursors are episodic rather than universal. Our results demonstrate the potential of SAI as a potential approach for precursor validation and underline the importance of multi-event, multi-parameter studies in moving the field toward reproducible earthquake forecasting science.</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

Evaluation of earthquake precursors using a multi-parameter Seismo-Atmospheric Anomaly Index

  • 期刊论文

  • 发表日期：

    2026年7月8日

查看全文
  • Zeeshan Haider,
  • Jianguo Yan,
  • Jean-Pierre Barriot

摘要

The search for reliable earthquake precursors remains one of the most challenging problems in geophysics. The lithosphere–atmosphere–ionosphere coupling (LAIC) mechanism has been widely proposed to explain reported pre-seismic anomalies in total electron content (TEC), Outgoing Long Wave Radiation (OLR), and land surface temperature (LST); however, the consistency and generalizability of these anomalies across various scenarios remain unclear. Here, we introduce a novel multi-parameter Seismo-Atmospheric Anomaly Index (SAI) that systematically integrates GNSS-based TEC, Swarm satellite TEC, MODIS-derived Land Surface Temperature (LST), and Outgoing Long Wave Radiation (OLR) into a single normalized score. A Nonlinear Auto Regressive with eXogenous inputs (NARX) machine learning model is employed to detect anomalies in OLR and LST. By incorporating a sliding 5-day time window, the SAI framework is explicitly designed to capture lagged or asynchronous anomalies across different atmospheric and ionospheric layers. Applying this method to five moderate-to-large Indonesian earthquakes (Mw 6.6–7.1, 2020–2023), we show that only the Mw 6.9 Bengkulu earthquake (18 August 2020) exhibited clear, multi-parameter precursors exceeding the anomaly detection threshold, while no significant anomalies were detected for the remaining four events. This dual outcome detecting positive anomalies in one case while ruling them out in others provides a rare quantitative evaluation of LAIC and highlights that these seismo-atmospheric precursors are episodic rather than universal. Our results demonstrate the potential of SAI as a potential approach for precursor validation and underline the importance of multi-event, multi-parameter studies in moving the field toward reproducible earthquake forecasting science.