Scholarlink

<p>Seismicity de-clustering involves categorizing seismic events in a catalog into mainshocks, aftershocks, foreshocks, and background events. Achieving accurate de-clustering is essential for interpreting seismic activity patterns and assessing geological risks. This study addresses the challenge of effectively separating these events in earthquake-prone regions through the development of a modified Multi-objective Adaptive Guided Differential Evolution (mMOAGDE) algorithm. The mMOAGDE algorithm introduces innovative features such as adaptive control parameters and three distinct mutation phases to improve diversity and exploration in the solution space. By utilizing Strength Pareto instead of crowding distance and incorporating an archive control mechanism to maintain a constant size of non-dominated solutions, the algorithm achieves a balance between exploration and exploitation. Furthermore, a binary version of mMOAGDE integrates logical adaptive guided exploration to enhance the separation of seismic events. The algorithm evaluates two objective functions, Global Moran’s Index (GMI) and Allan Factor (AF), in spatial and temporal domains to de-cluster seismic catalogs. Applied to 30 years of earthquake data from regions including Southern California, Indonesia, Iran, and Japan, the mMOAGDE is benchmarked against established algorithms such as ANSGA-III, CMOPSO, MOEA/D-UR, and FLEA. The results demonstrate that the de-clustered catalogs yield GMI values within the desired range (-1 to 1) and maximize AF values. Validation through cumulative plots, <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\lambda\)</EquationSource> </InlineEquation>-plots, Allan factor plots, and inter-event time versus inter-event distance plots confirms the algorithm’s effectiveness in distinguishing aftershocks from background seismicity. This research highlights the potential of the mMOAGDE algorithm as a powerful tool for seismicity de-clustering, offering significant insights into earthquake dynamics and enhancing seismic risk assessment methodologies.</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

Strength-Pareto Based Multi-objective Adaptive Guided Differential Evolution Algorithm for De-clustering Seismic Catalogs

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月24日

查看全文
  • Anuruddh Yadav,
  • Satyasai Jagannath Nanda,
  • Ashish Sharma

摘要

Seismicity de-clustering involves categorizing seismic events in a catalog into mainshocks, aftershocks, foreshocks, and background events. Achieving accurate de-clustering is essential for interpreting seismic activity patterns and assessing geological risks. This study addresses the challenge of effectively separating these events in earthquake-prone regions through the development of a modified Multi-objective Adaptive Guided Differential Evolution (mMOAGDE) algorithm. The mMOAGDE algorithm introduces innovative features such as adaptive control parameters and three distinct mutation phases to improve diversity and exploration in the solution space. By utilizing Strength Pareto instead of crowding distance and incorporating an archive control mechanism to maintain a constant size of non-dominated solutions, the algorithm achieves a balance between exploration and exploitation. Furthermore, a binary version of mMOAGDE integrates logical adaptive guided exploration to enhance the separation of seismic events. The algorithm evaluates two objective functions, Global Moran’s Index (GMI) and Allan Factor (AF), in spatial and temporal domains to de-cluster seismic catalogs. Applied to 30 years of earthquake data from regions including Southern California, Indonesia, Iran, and Japan, the mMOAGDE is benchmarked against established algorithms such as ANSGA-III, CMOPSO, MOEA/D-UR, and FLEA. The results demonstrate that the de-clustered catalogs yield GMI values within the desired range (-1 to 1) and maximize AF values. Validation through cumulative plots, \(\lambda\) -plots, Allan factor plots, and inter-event time versus inter-event distance plots confirms the algorithm’s effectiveness in distinguishing aftershocks from background seismicity. This research highlights the potential of the mMOAGDE algorithm as a powerful tool for seismicity de-clustering, offering significant insights into earthquake dynamics and enhancing seismic risk assessment methodologies.