Jun Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Jun Zhu has a total of 45 co-authors. Here are the top co-authors by number of shared publications:
- Zefeng Li
State Key Laboratory of Precision Geodesy University of Science and Technology of China Hefei China
7 shared publications - Xinming Wu
Laboratory of Seismology and Physics of the Earth’s Interior, School of Earth and Space Sciences, University of Science and Technology of China, Hefei, 230026, China
3 shared publications - Xu Si
School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, Georgia, U.S.A.
3 shared publications - Daoyuan Sun
State Key Laboratory of Precision Geodesy, School of Earth and Space Sciences, University of Science and Technology of China, Hefei, China
2 shared publications Ji Zhang
Laboratory of Seismology and Physics of Earth’s Interior, School of Earth and Space Science, University of Science and Technology of China , 230026 , Hefei, China
2 shared publications- Jiewen Li
Hubei Subsurface Multi‐scale Imaging Key Laboratory School of Geophysics and Geomatics China University of Geosciences Wuhan China
2 shared publications
Latest Publications
- Open Access
- 17 August 2026
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 247 (2)
Enhanced source discrimination between tectonic earthquakes and quarry blasts via a hybrid CNN–GNN trained with multiple stations
SUMMARY: Accurate seismic event discrimination has significant scientific and societal implications. While deep learning shows great promise for automatic feature extraction and classification, most existing models process seismograms from either a single station or a fixed number of stations. However, single-station approaches cannot exploit correlations between signals from different stations, while models that require a fixed number of stations lack the flexibility to adapt to networks with varying station counts. This study aims to enhance classification performance by utilizing seismograms from multiple stations while maintaining flexibility regarding the number of stations used. Each event is represented as a graph, where nodes correspond to stations with spectrogram features. To process these graph-structured data, we develop a hybrid model combining a convolutional neural network (CNN) and a graph neural network (GNN). The CNN first extracts features from individual spectrograms, and the GNN subsequently integrates these features across all nodes to perform the final classification. The model was trained on earthquakes and quarry explosions from Utah. We also compare our model’s performance with that of CNN classifiers, which employ post-processing strategies to derive the final event type. Results show that our model outperforms the CNN across all evaluation metrics. In summary, the model improves classification performance through effective multistation aggregation while also providing flexibility regarding the number of input stations.
- 17 August 2026
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 247 (2)
Open AccessEnhanced source discrimination between tectonic earthquakes and quarry blasts via a hybrid CNN–GNN trained with multiple stations
SUMMARY: Accurate seismic event discrimination has significant scientific and societal implications. While deep learning shows great promise for automatic feature extraction and classification, most existing models process seismograms from either a single station or a fixed number of stations. However, single-station approaches cannot exploit correlations between signals from different stations, while models that require a fixed number of stations lack the flexibility to adapt to networks with varying station counts. This study aims to enhance classification performance by utilizing seismograms from multiple stations while maintaining flexibility regarding the number of stations used. Each event is represented as a graph, where nodes correspond to stations with spectrogram features. To process these graph-structured data, we develop a hybrid model combining a convolutional neural network (CNN) and a graph neural network (GNN). The CNN first extracts features from individual spectrograms, and the GNN subsequently integrates these features across all nodes to perform the final classification. The model was trained on earthquakes and quarry explosions from Utah. We also compare our model’s performance with that of CNN classifiers, which employ post-processing strategies to derive the final event type. Results show that our model outperforms the CNN across all evaluation metrics. In summary, the model improves classification performance through effective multistation aggregation while also providing flexibility regarding the number of input stations.
- Open Access
- 28 April 2026
- Published by Elsevier in Earth and Planetary Science Letters
- Vol. 686
Near instantaneously triggered Mw 5.9 aftershock during the 2025 Mw 7.1 Dingri earthquake revealed by radar interferometry
- 28 April 2026
- Published by Elsevier in Earth and Planetary Science Letters
- Vol. 686
Open AccessNear instantaneously triggered Mw 5.9 aftershock during the 2025 Mw 7.1 Dingri earthquake revealed by radar interferometry
- Open Access
- 31 August 2025
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 52 (17)
Global Search of PKP Precursors With Graph Neural Network: Implications for Scatterers in the Lowermost Mantle
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 and estimate their strengths relative to global averages. Our exhaustive search produces unprecedentedly abundant scatterers, including new discoveries of those beneath Antarctica. The distribution of most global scatterers largely correlates with ultra‐low velocity zones (ULVZs). GraphCursor's potential extends to identifying other deep seismic phases and aids in promoting the discovery of new deep Earth structures.
- 31 August 2025
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 52 (17)
Open AccessGlobal Search of PKP Precursors With Graph Neural Network: Implications for Scatterers in the Lowermost Mantle
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 and estimate their strengths relative to global averages. Our exhaustive search produces unprecedentedly abundant scatterers, including new discoveries of those beneath Antarctica. The distribution of most global scatterers largely correlates with ultra‐low velocity zones (ULVZs). GraphCursor's potential extends to identifying other deep seismic phases and aids in promoting the discovery of new deep Earth structures.
- Open Access
- 29 March 2025
- Published by Wiley
Global Search of PKP Precursors with Graph Neural Network: Implications for Scatterers in the Lowermost mantle
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 developed GraphCursor, a model based on graph neural network, to detect PKP precursors across multiple stations. Trained on manually picked precursors, GraphCursor identified 30,429 precursor waveforms from 5,499 global earthquakes (2000–2024). We located scatterers by stacking multistation probability functions and estimated their strengths relative to global averages. Our exhaustive search produces unprecedentedly abundant scatterers, including new discoveries of those beneath North Pacific and Antarctica. Their distribution correlates with ultra-low velocity zones (ULVZs), highlighting connections between small- and large-scale structures in the lowermost mantle. GraphCursor’s potential extends to identifying other deep seismic phases and aids in promoting the discovery of new deep Earth structures.
- 29 March 2025
- Published by Wiley
Open AccessGlobal Search of PKP Precursors with Graph Neural Network: Implications for Scatterers in the Lowermost mantle
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 developed GraphCursor, a model based on graph neural network, to detect PKP precursors across multiple stations. Trained on manually picked precursors, GraphCursor identified 30,429 precursor waveforms from 5,499 global earthquakes (2000–2024). We located scatterers by stacking multistation probability functions and estimated their strengths relative to global averages. Our exhaustive search produces unprecedentedly abundant scatterers, including new discoveries of those beneath North Pacific and Antarctica. Their distribution correlates with ultra-low velocity zones (ULVZs), highlighting connections between small- and large-scale structures in the lowermost mantle. GraphCursor’s potential extends to identifying other deep seismic phases and aids in promoting the discovery of new deep Earth structures.
- 15 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 62 (01962892) , 1-13
SeisCLIP: A Seismology Foundation Model Pre-Trained by Multimodal Data for Multipurpose Seismic Feature Extraction
In seismology, while training a specific deep learning model for each task is common, it often faces challenges such as the scarcity of labeled data and limited regional generalization. Addressing these issues, we introduce SeisCLIP: a foundation model for seismology, leveraging contrastive learning during pretraining on multimodal data of seismic waveform spectra and the corresponding local and global event information. SeisCLIP consists of a transformer-based spectrum encoder and an multilayer perceptron (MLP)-based information encoder that are jointly pre-trained on massive data. During pretraining, contrastive learning aims to enhance representations by training two encoders to bring the corresponding waveform spectra and event information closer in the feature space, while distancing uncorrelated pairs. Remarkably, the pre-trained spectrum encoder offers versatile features, enabling its application across diverse tasks and regions. Thus, it requires only modest datasets for fine-tuning to specific downstream tasks. Our evaluations demonstrate SeisCLIP’s superior performance over baseline methods in tasks like event classification, localization, and focal mechanism analysis, even when using distinct datasets from various regions. In essence, SeisCLIP emerges as a promising foundational model for seismology, potentially revolutionizing foundation-model-based research in the domain.
- 15 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 62 (01962892) , 1-13
SeisCLIP: A Seismology Foundation Model Pre-Trained by Multimodal Data for Multipurpose Seismic Feature Extraction
In seismology, while training a specific deep learning model for each task is common, it often faces challenges such as the scarcity of labeled data and limited regional generalization. Addressing these issues, we introduce SeisCLIP: a foundation model for seismology, leveraging contrastive learning during pretraining on multimodal data of seismic waveform spectra and the corresponding local and global event information. SeisCLIP consists of a transformer-based spectrum encoder and an multilayer perceptron (MLP)-based information encoder that are jointly pre-trained on massive data. During pretraining, contrastive learning aims to enhance representations by training two encoders to bring the corresponding waveform spectra and event information closer in the feature space, while distancing uncorrelated pairs. Remarkably, the pre-trained spectrum encoder offers versatile features, enabling its application across diverse tasks and regions. Thus, it requires only modest datasets for fine-tuning to specific downstream tasks. Our evaluations demonstrate SeisCLIP’s superior performance over baseline methods in tasks like event classification, localization, and focal mechanism analysis, even when using distinct datasets from various regions. In essence, SeisCLIP emerges as a promising foundational model for seismology, potentially revolutionizing foundation-model-based research in the domain.
- Top Cited PapersOpen Access
- 6 January 2024
- Published by Springer Nature in Communications Earth & Environment
- Vol. 5 (1) , 1-13
An all-in-one seismic phase picking, location, and association network for multi-task multi-station earthquake monitoring
Earthquake monitoring is vital for understanding the physics of earthquakes and assessing seismic hazards. A standard monitoring workflow includes the interrelated and interdependent tasks of phase picking, association, and location. Although deep learning methods have been successfully applied to earthquake monitoring, they mostly address the tasks separately and ignore the geographic relationships among stations. Here, we propose a graph neural network that operates directly on multi-station seismic data and achieves simultaneous phase picking, association, and location. Particularly, the inter-station and inter-task physical relationships are informed in the network architecture to promote accuracy, interpretability, and physical consistency among cross-station and cross-task predictions. When applied to data from the Ridgecrest region and Japan, this method showed superior performance over previous deep learning-based phase-picking and localization methods. Overall, our study provides a prototype self-consistent all-in-one system of simultaneous seismic phase picking, association, and location, which has the potential for next-generation automated earthquake monitoring.
- 6 January 2024
- Published by Springer Nature in Communications Earth & Environment
- Vol. 5 (1) , 1-13
Top Cited PapersOpen AccessAn all-in-one seismic phase picking, location, and association network for multi-task multi-station earthquake monitoring
Earthquake monitoring is vital for understanding the physics of earthquakes and assessing seismic hazards. A standard monitoring workflow includes the interrelated and interdependent tasks of phase picking, association, and location. Although deep learning methods have been successfully applied to earthquake monitoring, they mostly address the tasks separately and ignore the geographic relationships among stations. Here, we propose a graph neural network that operates directly on multi-station seismic data and achieves simultaneous phase picking, association, and location. Particularly, the inter-station and inter-task physical relationships are informed in the network architecture to promote accuracy, interpretability, and physical consistency among cross-station and cross-task predictions. When applied to data from the Ridgecrest region and Japan, this method showed superior performance over previous deep learning-based phase-picking and localization methods. Overall, our study provides a prototype self-consistent all-in-one system of simultaneous seismic phase picking, association, and location, which has the potential for next-generation automated earthquake monitoring.
- Open Access
- 28 November 2023
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 236 (2) , 979-993
Deep learning and transfer learning of earthquake and quarry-blast discrimination: applications to southern California and eastern Kentucky
Discrimination between tectonic earthquakes (EQs) and quarry blasts is important for accurate EQ cataloguing and seismic hazard analysis. However, reliable classification of these two types of seismic events is challenging with no prior knowledge of source parameters. Here, we applied deep learning to perform this classification task in southern California and eastern Kentucky. Since the two regions differ significantly in available labelled data, class imbalance and waveform characteristics, we adopted different strategies for them. We directly trained a convolutional neural network (CNN) for southern California due to its data abundancy. To alleviate the class imbalance, the blast data were augmented by repeated sampling. The model for California yields F1-scores of >83.5 per cent when estimated by individual stations and >98.1 per cent by network average (i.e. averaging the CNN’s outputs on all available stations for each event). As eastern Kentucky has a much smaller data size, we apply transfer learning to the pre-trained California model to fit the Kentucky data. The transfer-learned model yields F1-scores of >86.9 per cent when estimated by individual stations and >96.7 per cent by network average. The transfer-learned model outperforms the model re-trained from scratch for the Kentucky data. Gradient-weighted class activation mapping shows the S onset and the S long-period coda are important to identify EQs and blasts, respectively. By visual inspections of the seismograms, the source locations, the origin time and the P-wave polarities, we verified that most of the events falsely predicted by our models are actually mislabelled by seismic analysts. Our models thus show great potential in helping seismic analysts find those mislabelled events which remain hidden in the historical catalogue. Our results demonstrate that deep learning can achieve high accuracy in seismic event discrimination and that transfer learning is effective and efficient to generalize deep learning models across different regions.
- 28 November 2023
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 236 (2) , 979-993
Open AccessDeep learning and transfer learning of earthquake and quarry-blast discrimination: applications to southern California and eastern Kentucky
Discrimination between tectonic earthquakes (EQs) and quarry blasts is important for accurate EQ cataloguing and seismic hazard analysis. However, reliable classification of these two types of seismic events is challenging with no prior knowledge of source parameters. Here, we applied deep learning to perform this classification task in southern California and eastern Kentucky. Since the two regions differ significantly in available labelled data, class imbalance and waveform characteristics, we adopted different strategies for them. We directly trained a convolutional neural network (CNN) for southern California due to its data abundancy. To alleviate the class imbalance, the blast data were augmented by repeated sampling. The model for California yields F1-scores of >83.5 per cent when estimated by individual stations and >98.1 per cent by network average (i.e. averaging the CNN’s outputs on all available stations for each event). As eastern Kentucky has a much smaller data size, we apply transfer learning to the pre-trained California model to fit the Kentucky data. The transfer-learned model yields F1-scores of >86.9 per cent when estimated by individual stations and >96.7 per cent by network average. The transfer-learned model outperforms the model re-trained from scratch for the Kentucky data. Gradient-weighted class activation mapping shows the S onset and the S long-period coda are important to identify EQs and blasts, respectively. By visual inspections of the seismograms, the source locations, the origin time and the P-wave polarities, we verified that most of the events falsely predicted by our models are actually mislabelled by seismic analysts. Our models thus show great potential in helping seismic analysts find those mislabelled events which remain hidden in the historical catalogue. Our results demonstrate that deep learning can achieve high accuracy in seismic event discrimination and that transfer learning is effective and efficient to generalize deep learning models across different regions.
- Open Access
- 31 October 2023
- Published by Springer Nature in Research Square
Multi-task multi-station earthquake monitoring: An all-in-one seismic Phase picking, Location, and Association Network (PLAN)
Earthquake monitoring is vital for understanding the physics of earthquakes and assessing seismic hazards. A standard monitoring workflow includes the interrelated and interdependent tasks of phase picking, association, and location. Although deep learning methods have been successfully applied to earthquake monitoring, they mostly address the tasks separately and ignore the geographic relationships among stations. Here, we propose a graph neural network that operates directly on multi-station seismic data and achieves simultaneous phase picking, association, and location. Particularly, the inter-station and inter-task physical relationships are informed in the network architecture to promote accuracy, interpretability, and physical consistency among cross-station and cross-task predictions. When applied to data from the Ridgecrest region and Japan regions, this method showed superior performance over previous deep learning-based phase-picking and localization methods. Overall, our study provides for the first time a prototype self-consistent all-in-one system of simultaneous seismic phase picking, association, and location, which has the potential for next-generation automated earthquake monitoring.
- 31 October 2023
- Published by Springer Nature in Research Square
Open AccessMulti-task multi-station earthquake monitoring: An all-in-one seismic Phase picking, Location, and Association Network (PLAN)
Earthquake monitoring is vital for understanding the physics of earthquakes and assessing seismic hazards. A standard monitoring workflow includes the interrelated and interdependent tasks of phase picking, association, and location. Although deep learning methods have been successfully applied to earthquake monitoring, they mostly address the tasks separately and ignore the geographic relationships among stations. Here, we propose a graph neural network that operates directly on multi-station seismic data and achieves simultaneous phase picking, association, and location. Particularly, the inter-station and inter-task physical relationships are informed in the network architecture to promote accuracy, interpretability, and physical consistency among cross-station and cross-task predictions. When applied to data from the Ridgecrest region and Japan regions, this method showed superior performance over previous deep learning-based phase-picking and localization methods. Overall, our study provides for the first time a prototype self-consistent all-in-one system of simultaneous seismic phase picking, association, and location, which has the potential for next-generation automated earthquake monitoring.
- Open Access
Accelerated Linearized Laplace Approximation for Bayesian Deep Learning
Laplace approximation (LA) and its linearized variant (LLA) enable effortless adaptation of pretrained deep neural networks to Bayesian neural networks. The generalized Gauss-Newton (GGN) approximation is typically introduced to improve their tractability. However, LA and LLA are still confronted with non-trivial inefficiency issues and should rely on Kronecker-factored, diagonal, or even last-layer approximate GGN matrices in practical use. These approximations are likely to harm the fidelity of learning outcomes. To tackle this issue, inspired by the connections between LLA and neural tangent kernels (NTKs), we develop a Nystrom approximation to NTKs to accelerate LLA. Our method benefits from the capability of popular deep learning libraries for forward mode automatic differentiation, and enjoys reassuring theoretical guarantees. Extensive studies reflect the merits of the proposed method in aspects of both scalability and performance. Our method can even scale up to architectures like vision transformers. We also offer valuable ablation studies to diagnose our method. Code is available at \url{https://github.com/thudzj/ELLA}.
Open AccessAccelerated Linearized Laplace Approximation for Bayesian Deep Learning
Laplace approximation (LA) and its linearized variant (LLA) enable effortless adaptation of pretrained deep neural networks to Bayesian neural networks. The generalized Gauss-Newton (GGN) approximation is typically introduced to improve their tractability. However, LA and LLA are still confronted with non-trivial inefficiency issues and should rely on Kronecker-factored, diagonal, or even last-layer approximate GGN matrices in practical use. These approximations are likely to harm the fidelity of learning outcomes. To tackle this issue, inspired by the connections between LLA and neural tangent kernels (NTKs), we develop a Nystrom approximation to NTKs to accelerate LLA. Our method benefits from the capability of popular deep learning libraries for forward mode automatic differentiation, and enjoys reassuring theoretical guarantees. Extensive studies reflect the merits of the proposed method in aspects of both scalability and performance. Our method can even scale up to architectures like vision transformers. We also offer valuable ablation studies to diagnose our method. Code is available at \url{https://github.com/thudzj/ELLA}.
- Open Access
Regret Analysis for Hierarchical Experts Bandit Problem
We study an extension of standard bandit problem in which there are R layers of experts. Multi-layered experts make selections layer by layer and only the experts in the last layer can play arms. The goal of the learning policy is to minimize the total regret in this hierarchical experts setting. We first analyze the case that total regret grows linearly with the number of layers. Then we focus on the case that all experts are playing Upper Confidence Bound (UCB) strategy and give several sub-linear upper bounds for different circumstances. Finally, we design some experiments to help the regret analysis for the general case of hierarchical UCB structure and show the practical significance of our theoretical results. This article gives many insights about reasonable hierarchical decision structure.
Open AccessRegret Analysis for Hierarchical Experts Bandit Problem
We study an extension of standard bandit problem in which there are R layers of experts. Multi-layered experts make selections layer by layer and only the experts in the last layer can play arms. The goal of the learning policy is to minimize the total regret in this hierarchical experts setting. We first analyze the case that total regret grows linearly with the number of layers. Then we focus on the case that all experts are playing Upper Confidence Bound (UCB) strategy and give several sub-linear upper bounds for different circumstances. Finally, we design some experiments to help the regret analysis for the general case of hierarchical UCB structure and show the practical significance of our theoretical results. This article gives many insights about reasonable hierarchical decision structure.