Global Search of PKP Precursors With Graph Neural Network: Implications for Scatterers in the Lowermost Mantle

DOI number:10.1029/2025gl115952

Affiliation of Author(s):中国科学技术大学地球和空间科学学院

Journal:Geophysical Research Letters

Place of Publication:美国

Abstract:Seismic waves preceding the core phase PKIKP, known as PKP precursors, have been used to map small scatterers near the core‐mantle boundary (CMB). However, their low signal‐to‐noise ratio has required manual identification, hindering systematic and comprehensive analysis. Here, we develop GraphCursor, a model based on Graph Neural Networks (GNNs), to detect PKP precursors across multiple stations. Trained on manually picked precursors, GraphCursor identifies 37,512 precursor waveforms from 5,499 global earthquakes (2000–2024). We locate scatterers by stacking multistation probability functions

First Author:Shang Ma

Co-author:Zefeng Li, Daoyuan Sun, Yulong Su, Jiewen Li, Xu Si, Jun Zhu

Indexed by:Journal paper

Correspondence Author:Zefeng Li

Document Code:43

Discipline:Natural Science

First-Level Discipline:Geophysics

Document Type:SCI

Volume:52

Issue:17

Page Number:e2025GL115952

ISSN No.:0094-8276

Translation or Not:no

Date of Publication:2025-08-31

Links to published journals:https://doi.org/10.1029/2025gl115952