Monitoring how stress accumulates and evolves toward failure is a central challenge across many scientific and engineering disciplines. In aerospace engineering, civil infrastructure, mining, nuclear power, and materials science, stress–strain trajectories are routinely monitored to assess structural integrity, safety margins, and remaining lifetime. These measurements are typically obtained using active approaches—such as ultrasonic probing, controlled vibration tests, strain gauges, or periodic inspections—which are often expensive, intrusive, spatially limited, and difficult to maintain continuously.
My current research introduces a fundamentally different approach to this problem: a stress-sensitive frequency-domain transform (SSf) that enables continuous, passive monitoring of stress evolution using only ambient acoustic or seismic noise.
SSf is based on the discovery that stress-induced changes in solids manifest as systematic and repeatable redistributions of energy between nearby frequency bands in recorded wavefields. Rather than tracking absolute amplitudes or travel times, SSf focuses on relative energy transfer across adjacent frequencies, producing a low-dimensional representation that evolves smoothly with loading and deformation.
This design allows SSf to extract stress-sensitive information directly from background signals, without requiring active sources, controlled excitation, or identifiable wave arrivals. As a result, SSf can operate continuously and non-intrusively, even in complex or noisy environments.
In engineering and industrial contexts—including aerospace structures, bridges and dams, tunnels and mines, nuclear power plant components, and other critical infrastructure—SSf offers two immediate advantages. First, it eliminates the need for active excitation, reducing operational cost and complexity. Second, it provides an independent, physics-based stress indicator that can complement existing sensor systems and enable continuous verification of structural state.
While stress monitoring is well established in engineered systems, earthquake science has long lacked an equivalent capability. In tectonic environments, controlled active sources cannot be deployed repeatedly or at depth, and fault-zone stress evolution has remained largely unobservable between earthquakes.
Because SSf relies only on ambient seismic noise, it can be applied wherever continuous seismic recordings exist. Using SSf, I have identified consistent stress-related trajectories associated with slow-slip cycles, induced seismicity, and large tectonic earthquakes, including the 2011 Tohoku earthquake, 2010 Maule earthquake, 2002 Denali earthquake, and the 2023 Turkey–Syria earthquake. Across these cases, SSf reveals coherent loading, failure, and post-event relaxation behavior that is not captured by conventional seismic metrics such as amplitude, spectral power, or event rates.
Preprint available (arXiv:2509.00268)
Github: https://github.com/Naderss/earthquake-precursors-ssf
U.S. Patent Pending: Application No. 63/870,020, “Methods and Systems for Detecting Earthquake Precursors via Stress-Sensitive Transformations of Seismic Noise”, filed on August 25, 2025.
The Matrix Profile:
I am proud to be one of the original contributors to the development of the Matrix Profile, a powerful framework for time series data mining and analysis. The Matrix Profile enables efficient detection of patterns, motifs, and anomalies in large datasets, transforming the way complex time series data are explored and interpreted. This work has opened new avenues for research and practical applications across diverse domains, with particular impact on seismic data analysis and other time series data mining fields (link 1, link 2).
Publications:
Peter M. Shearer, Nader Shakibay Senobari; Continuous Aftershock Hum for over Ten Days Following the 2019 Ridgecrest, California, Earthquakes Observed with Borehole Seismometers. Seismological Research Letters 2025; doi: https://doi.org/10.1785/0220250017
Nader Shakibay Senobari, Peter Sherear, Gareth J Funning, Ya Zhu, Zachary Zimmerman, Phillip Brisk, Eamonn Keogh, 2024, "The matrix profile in seismology: template matching of everything with everything", Journal of Geophysical Research: Solid Earth 129, no. 2 (2024): e2023JB027122.
Peter Shearer, Nader Shakibay Senobari, Yuri Fialko. Implications of a reverse polarity earthquake pair on fault friction and stress heterogeneity near Ridgecrest, California, Journal of Geophysical Research: Solid Earth, Accepted, 2024
Ryan Mercer, Sara Alaee, Alireza Abdoli, Nader Shakibay Senobari, Shailendra Singh, Amy C. Murillo, Eamonn J. Keogh. Introducing the contrast profile: a novel time series primitive that allows real world classification.: Data Min. Knowl. Discov. 36(2): 877-915, 2022.
Zhu, Yan, Shaghayegh Gharghabi, Diego Furtado Silva, Hoang Anh Dau, Chin-Chia Michael Yeh, Nader Shakibay Senobari, Abdulaziz Almaslukh, Kaveh Kamgar, Zachary Zimmerman, Gareth Funning, Abdullah Mueen,Eamonn Keogh. "The Swiss Army Knife of Time Series Data Mining: Ten Useful Things you can do with the Matrix Profile and Ten Lines of Code."Data Mining and Knowledge Discovery 2020[ pdf].
Zimmerman, Zachary, Kaveh Kamgar, Nader Shakibay Senobari, Brian Crites, Gareth Funning, Philip Brisk, and Eamonn Keogh. "Matrix Profile XIV: Scaling Time Series Motif Discovery with GPUs to Break a Quintillion Pairwise Comparisons a Day and Beyond." In Proceedings of the ACM Symposium on Cloud Computing, pp. 74-86. 2019.
Zimmerman, Zachary, Nader Shakibay Senobari, Gareth Funning, Evangelos Papalexakis, Samet Oymak, Philip Brisk, and Eamonn Keogh. "Matrix Profile XVIII: Time Series Mining in the Face of Fast Moving Streams using a Learned Approximate Matrix Profile." IEEE ICDM 2019. [pdf]
Madrid, Frank, Shima Imani, Ryan Mercer, Zachary Zimmerman, Nader Shakibay Senobari, and Eamonn Keogh. "Matrix Profile XX: Finding and Visualizing Time Series Motifs of All Lengths using the Matrix Profile." In 2019 IEEE International Conference on Big Knowledge (ICBK), pp. 175-182. IEEE, 2019. IEEE Big Knowledge 2019. [pdf]
Yan Zhu, Zachary Zimmerman, Nader Shakibay Senobari, Chin-Chia M. Yeh, Gareth Funning, Abdullah Mueen, Philip Brisk, and Eamonn Keogh, "Exploiting a Novel Algorithm and GPUs to Break the Ten Quadrillion Pairwise Comparisons Barrier for Time Series Motifs and Joins," KAIS Journal, 2017.
Yan Zhu, Zachary Zimmerman, Nader Shakibay Senobari, Chin-Chia M. Yeh, Gareth Funning, Abdullah Mueen, Philip Brisk, and Eamonn Keogh, "Matrix Profile II: Exploiting a Novel Algorithm and GPUs to Break the One Hundred Million Barrier for Time Series Motifs and Joins," IEEE International Conference on Data Mining (ICDM), 2016.
Data set: [link]
SEC-C (Super-Efficient Cross-Correlation) is a scalable algorithm for detecting repeating and weakly similar patterns in massive time-series datasets. The method enables near-exhaustive cross-correlation analysis at a computational cost that is orders of magnitude lower than classical approaches, making it practical for continuous, long-duration recordings.
Originally developed in the context of seismology, SEC-C allows efficient identification of repeating earthquakes, low-magnitude events, and subtle waveform similarities embedded in noisy data. More broadly, the framework provides a general solution for large-scale similarity search in time-series, with applications spanning geophysics, acoustics, structural monitoring, and other sensor-driven domains.
The accompanying paper introduces the theoretical foundations of SEC-C, demonstrates its computational advantages, and validates its performance on real seismic datasets, establishing SEC-C as a practical alternative to traditional cross-correlation pipelines for modern, data-intensive science.
Publication:
Nader Shakibay Senobari, G J Funning, E Keogh, Y Zhu, C-C M Yeh, Z Zimmerman, and A Mueen, 2018, Super-Efficient Cross-Correlation (SEC-C): A fast matched filtering code suitable for desktop computers, Seismol. Res. Lett., doi:10.1785/0220180122
Github: [link]
This work introduces a fully automated framework developed to overcome a fundamental limitation of traditional repeating earthquake detection methods, which typically require dense, high-quality seismic networks such as those in central California or Japan. In contrast, the proposed approach was designed to operate reliably with sparse and imperfect seismic networks in both space and time.
The method was successfully applied to regions such as the San Francisco Bay Area, where network coverage has historically been limited. Repeating earthquakes—commonly interpreted as signatures of aseismic fault creep—identified by this framework were consistent with independent InSAR observations, correctly delineating known creeping fault zones while also suggesting previously unrecognized creeping segments. Notably, a newly identified creeping segment reported in this study was later investigated through fieldwork by geologists from the United States Geological Survey, who confirmed geological evidence of fault creep in the region.
Publication:
Nader Shakibay Senobari, and G. J. Funning, Widespread fault creep in the northern San Francisco Bay Area revealed by multi-station cluster detection of repeating earthquakes, Geophys. Res. Lett., in press, 2019, doi:10.1029/2019GL082766. (Preprint available at Earth and Space Science Open Archive, doi: 10.1002/essoar.10500868.1)
Sugan, Monica, Stefano Campanella, Alessandro Vuan, and Nader Shakibay Senobari. "A Python Code for Detecting True Repeating Earthquakes from Self‐Similar Waveforms (FINDRES)." Seismological Society of America 93, no. 5 (2022): 2847-2857.
Github: [link]
Interferometric Synthetic Aperture Radar (InSAR) & Seismology Integration
Developed novel methodologies integrating InSAR with long-period teleseismic waveform analysis to improve earthquake source characterization and Earth velocity models. Demonstrated that InSAR-derived centroids can be used to identify and correct systematic travel-time biases in global seismic inversions, leading to improved event locations and source mechanisms. Extensive experience with Sentinel-1 and ERS InSAR processing, elastic dislocation modeling, seismic waveform simulation (SPECFEM3D), and large-scale geophysical data integration, with applications to hazard assessment and Earth structure characterization.
Current Projects:
A Supremely Accurate and Remarkably Efficient Earthquake Detection and Seismic Quality Monitoring System (paper in prep)
A method for all-length motif discovery using the Matrix Profile has been developed, and the corresponding paper is currently in preparation