I work at the intersection of geophysics, statistical modeling, and machine learning — applying deep learning, signal processing, and probabilistic methods to problems in seismology, geosciences, wireless signals, and beyond.
📍 Pocatello, ID | 📧 danedwells@gmail.com | 🔗 LinkedIn | Website
Languages
ML & Data Science
Web & Frameworks
Design & GIS
Specializations
- Deep Learning: CNNs, Transformers, GANs, Autoencoders, Physics-Informed Neural Networks
- Geophysics: Seismology, Ambient Noise Tomography, Seismic Phase Picking, Earthquake Location
- Signal Processing: Time-frequency analysis, Wavelet methods, IQ data, RF signal classification
- Probabilistic Methods: Bayesian inference, Hamiltonian Monte Carlo, Inversion Theory
- HPC: SLURM, GPU acceleration, parallelization
Improving earthquake early warning by implementing static and dynamic Bayesian priors and context-driven GRU architectures to forecast and compute magnitude, location, and timing of earthquakes.
Flexible framework for accessible, automatic application of AI and ML to scientific data. Features auto-encoders and dimensionality reduction wrappers for rapid deployment across diverse input types.
Supervised ML for monazite classification; addressing data scarcity challenges in REE geoscience.
Physics-informed neural networks to optimize subsurface serpentinization for clean hydrogen production.
ML pipeline for portable detection, classification, and open-set recognition of unknown wireless signals using GANs, CNNs, and transformer architectures. Deployed on field hardware.
Peer-Reviewed Journals
-
Wells, D. E. et al. (2024). Wasatch Fault Structure from Machine Learning Arrival Times and High-Precision Earthquake Locations. Bulletin of the Seismological Society of America. DOI
-
Wells, D., Lin, F. C., Pankow, K., et al. (2022). Combining dense seismic arrays and broadband data to image subsurface velocity structure in geothermally active South-central Utah. Journal of Geophysical Research: Solid Earth. DOI
-
Pankow, K. L., ... Wells, D., et al. (2020). Responding to the 2020 Magna, Utah earthquake sequence during COVID-19. Seismological Research Letters. DOI
-
Wei, X., Saha, D., Wells, D., & Quach, A. (2025). Contrasting Time-Frequency Representations for Unknown Waveform Detection. IEEE ICC 2025.
In Preparation
- Wells, D. et al. (2025). A Comparative Analysis of Feature Extraction and Distance Loss Functions for Unknown Waveform Detection. IEEE Transactions on Signal Processing.
- Wells, D. E. et al. (2025). Evidence for interaction of Wasatch Fault segments at depth. Seismica.
| Year | Role | Funder | Project | Amount |
|---|---|---|---|---|
| 2025 | PI | NNSA (NA-241) | Seismic Sensor Infrastructure at WIPP | $130,000 |
| 2024 | Co-I | INL LDRD | GREENR — Rare Earth Element Geometallurgy | $688,000 |
| 2023 | Fellow | USGS Mendenhall | Machine Learning for Global Seismology | ~$300,000 |
US202250192904A1— Spectrum Monitoring and Analysis (WiFIRE)US11251889B2— Wireless Signal Monitoring and Analysis (WiFIRE)US12418349B2— Spectrum Monitoring and Analysis (WiFIRE)
| Platform | Link |
|---|---|
| danedwells@gmail.com | |
| dan-wells-591b49a8 |