Open research questions in Seismology and Earthquake Studies

What the literature leaves open

• While the method identifies time-frequency features in the 30-40 Hz range for foreshocks, the physical mechanisms explaining why these specific frequencies are indicative of foreshock activity remain to be fully established.

  Analysis of Foreshocks and Aftershocks in a microseismic sequence in Switzerland using Explainable AI · 2026 · DOI

• The spatial distribution of seismic stations is often sparse and non-uniform across the geospatial domain. The model might exhibit artifacts in producing FAS maps for areas lacking measurement data. Focal mechanisms are not included as conditional variables due to the concentration of earthquake data around the Hayward and San Andreas Faults.

  Learning earthquake ground motions via conditional generative modeling · 2026 · DOI

• The approach is limited to a short interval after the P-wave onset. The use of pre-trained vision models may not be optimal for seismic signal classification. The results are based on a specific dataset and may not generalize to other datasets.

  Real-time seismic signal classification using deep vision architectures for earthquake early warning systems · 2026 · DOI

• First, more sophisticated but computationally more demanding optimization 12 schemes, such as stochastic gradient descent (SGD) or SGD with momentum (Bottou, 2012), could be explored.

  Effectively Distinguishing Blast and Earthquake Sources in Eastern Canada · 2026 · DOI

• The lack of efficient processing strategies capable of handling high data complexity both in terms of size and signal quality. The need for a near real-time selector of high-quality DAS channels.

  Near real-time channel selection for Distributed Acoustic Sensing technology · 2026 · DOI

• The earthquake highlighted the need for effective communication and coordination between government agencies and emergency responders. The earthquake demonstrated the importance of rapid functional recovery strategies.

  The 2024 Hualien, Taiwan earthquake: NZSEE learning from earthquakes reconnaissance report · 2026 · DOI

• The paper identifies a need for further research on seismic resilience developments in Taiwan. The paper highlights the importance of understanding the impacts of earthquakes on infrastructure and buildings.

  The 2024 Hualien, Taiwan earthquake: NZSEE learning from earthquakes reconnaissance report · 2026 · DOI

• Further studies can focus on refining the seismotectonic source model and attenuation relationships. The use of other seismic hazard assessment techniques, such as machine learning algorithms, can be explored.

  An updated seismic hazards and seismotectonic studies for Sinai Peninsula, Egypt · 2026 · DOI

• There is a need to evaluate seismic hazards in the Sinai Peninsula using probabilistic and deterministic techniques. The region's complex tectonic setting and significant economic resources make it a critical area for seismic hazard assessment.

  An updated seismic hazards and seismotectonic studies for Sinai Peninsula, Egypt · 2026 · DOI

• Future work will focus on three directions: introducing richer structural metadata such as three-dimensional sensor coordinates, embedment depth, reinforcement layout, structural zoning, and finite-element connectivity; adding calibrated uncertainty intervals through conformal, quantile, or Bayesian forecasting; and validating the framework on additional dams, construction stages, and hydrological conditions.

  A Multi-Task Temporal Fusion Framework for 48 h Ahead Joint Prediction of Dam Crack Responses and Rebar Stress from Multi-Source Monitoring Data · 2026 · DOI

• The complexity of tremor signals, which can be difficult to capture using conventional earthquake location techniques. The need to quantify aleatoric uncertainty in tremor source localization, which can be challenging due to the limited availability of seismic data.

  Single station seismic observations enable high resolution localization of tectonic tremor sources using a vision transformer · 2026 · DOI

• Traditional stationary methods are insufficient for accurate forecasting of seismic risks. The integration of non-stationarity in disaster risk forecasting is necessary for accurate modeling.

  Integration of Non-Stationarity in Disaster Risk Forecasting: A Multidisciplinary Stochastic Analysis of Seismic Dynamics in Azerbaijan based on Hidden Markov Models · 2026 · DOI

• The lack of a workflow for compiling and validating heterogeneous hazard data.

  Natural hazard events dataset for the Garrotxa Region (Catalonia, Spain): a foundational step toward multi-hazard risk assessment · 2026 · DOI

• There is a need for rigorous testing and evaluation of earthquake forecasting models against prospective data and benchmarking. Prior work has focused on developing and evaluating individual models, but there is a lack of comprehensive evaluation and benchmarking of multiple models.

  Select earthquake forecasting models demonstrate consistency with prospective decadal observations in California · 2026 · DOI

• Ongoing developments include spectral moment magnitudes and DAS strain-rate-based magnitude estimates to improve characterization of small-magnitude events.

  Multi-Scale On-Fault Seismic Monitoring at the Bedretto Underground Laboratory: An Operational Framework · 2026 · DOI

• This study explored the integration of spatial and non-spatial data to design an intelligent, location-aware dashboard that supports proactive earthquake disaster management.

  Integration of Spatial Data from Heterogeneous Sources and Their Handling in a Dashboard for Earthquake Disaster Management · 2026 · DOI

• Investigating the application of the framework to other fields where temporal structure is important.

  Self-supervised seismogram learning using Siamese CNN · 2026 · DOI

• The operational efficacy of real-time kinematic positioning remains sensitive to signal obstruction. Current early warning systems rely on fragmented, static methodologies.

  Geological Hazard Prediction and Prevention: A Review of Mechanisms, Monitoring, and Mitigation · 2026 · DOI

• Contemporary predictive models are bifurcated between deterministic physical models and data-driven artificial intelligence algorithms.

  Geological Hazard Prediction and Prevention: A Review of Mechanisms, Monitoring, and Mitigation · 2026 · DOI

• Limited implementation of multi-hazard perspectives at local scales.

  Natural hazard events dataset for the Garrotxa Region (Catalonia, Spain): a foundational step toward multi-hazard risk assessment · 2026 · DOI

• To improve the model's performance in the low-frequency and high-frequency parts. To evaluate the model's performance using different datasets and conditional variables. To develop more advanced ground motion prediction methodologies.

  Learning earthquake ground motions via conditional generative modeling · 2026 · DOI

• Investigating the use of other deep learning architectures for seismic signal classification. Evaluating the effectiveness of the proposed approach for other types of seismic signals. Investigating the use of transfer learning for seismic signal classification.

  Real-time seismic signal classification using deep vision architectures for earthquake early warning systems · 2026 · DOI

• Further analysis of the seismicity in the Gargano area using the new catalog. Application of the CASP software to other seismic networks to improve earthquake detection capabilities. Investigation of the seismotectonic and seismic hazard characteristics in the Gargano area.

  The new seismic catalog of the Gargano area (Southern Italy) after a decade of seismic monitoring by OTRIONS network · 2026 · DOI

• Further testing of ORION on different datasets and hardware configurations. Exploration of the application of ORION to other types of seismic data.

  Near real-time channel selection for Distributed Acoustic Sensing technology · 2026 · DOI

• Further evaluation of the framework's performance on different datasets and regions, - Investigation of the potential for shorter time windows, - Development of methods to improve location accuracy and reduce…

  Single station seismic observations enable high resolution localization of tectonic tremor sources using a vision transformer · 2026 · DOI

Most-cited papers in Seismology and Earthquake Studies

• Deep Learning for Geophysics: Current and Future Trends · Reviews of Geophysics · 2021 · 554 citations
• Deep learning for geological hazards analysis: Data, models, applications, and opportunities · Earth-Science Reviews · 2021 · 307 citations
• Recent Advances in Internet of Things Solutions for Early Warning Systems: A Review · Sensors · 2022 · 181 citations
• Earthquake Phase Association Using a Bayesian Gaussian Mixture Model · Journal of Geophysical Research Solid Earth · 2022 · 175 citations
• SeisBench—A Toolbox for Machine Learning in Seismology · Seismological Research Letters · 2022 · 169 citations
• LOC-FLOW: An End-to-End Machine Learning-Based High-Precision Earthquake Location Workflow · Seismological Research Letters · 2022 · 155 citations
• An overview of the Illinois Basin – Decatur Project · Greenhouse Gases Science and Technology · 2014 · 140 citations
• Machine learning in microseismic monitoring · Earth-Science Reviews · 2023 · 137 citations
• Feasibility of Stochastic Gradient Boosting Approach for Evaluating Seismic Liquefaction Potential Based on SPT and CPT Case Histories · Journal of Performance of Constructed Facilities · 2019 · 133 citations
• Automated identification, location, and volume estimation of rockfalls at Piton de la Fournaise volcano · Journal of Geophysical Research Earth Surface · 2014 · 121 citations

Most recent work

• CDU-Net: An adaptive conditional denoising model for microseismic signals in complex environments · Underground Space · 2026
• Zero-Shot Vision-Language Model for Rapid Damaged Bridge Extraction in Emergency Response: A Case Study of the 2025 Myanmar Earthquake · IEEE Geoscience and Remote Sensing Letters · 2026
• Earthquake Magnitude Detection Utilizing a Novel Hybrid Earth–Transformer–LSTM Architecture · Future Internet · 2026
• FEDONet: Fourier-embedded DeepONet for spectrally accurate operator learning · Journal of Computational Physics · 2026
• Improvement of seismic wave detection using genetic algorithm–based optimized FIR filters · Soil Dynamics and Earthquake Engineering · 2026
• Location for microseismic event clusters in deep tunnels based on coda wave interference · Tunnelling and Underground Space Technology · 2026
• A Bi-LSTM Attention Mechanism for Monitoring Seismic Events—Solving the Issue of Noise & Interpretability · Technologies · 2026
• A physics-informed deep learning and probabilistic inference framework for real-time single-station earthquake detection and magnitude estimation · Bulletin of Earthquake Engineering · 2026
• Learning earthquake ground motions via conditional generative modeling · Nature Communications · 2026
• Machine Learning–Based Automatic Microseismic Event Detection During the 17 August 2015 Mw 4.6 Induced Earthquake Sequence in Northern Montney, British Columbia, Canada · Bulletin of the Seismological Society of America · 2026

Related topics in Computer Science