About
Pawan Bharadwaj works at the interface of seismology, inverse theory, signal processing, and scientific machine learning. His research develops algorithms that recover useful physical information from complex seismic wavefields, including earthquake-source signatures, subsurface structure, and weak signals hidden by scattering and noise. Current work emphasizes physics-guided neural architectures for earthquake source characterization, seismic phase separation, passive seismic imaging, tomography, and time-lapse monitoring applications.
Research Interests
Education
Postdoctoral Associate (Applied Mathematics)
Massachusetts Institute of Technology (MIT), Cambridge, USA.
PhD (Geophysics), 2016
Delft University of Technology (TU Delft), Delft, The Netherlands.
Master of Science and Technology (5-year Integrated, Applied Geophysics), 2012
Indian Institute of Technology (IIT), Dhanbad, India.
Recent Publications
- Learning Earthquake Sources Using Symmetric Autoencoders — Bulletin of the Seismological Society of America (2025)
- Symmetric autoencoders for retrieving void-generated ringing modes and backscatter from vehicle seismic noise — Geophysics (2025)
- Coherent Spectral Feature Extraction Using Symmetric Autoencoders — IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (2025)
- Redshift of earthquakes via focused blind deconvolution of teleseisms — Geophysical Journal International (2020)
- Focused Blind Deconvolution — IEEE Transactions on Signal Processing (2019)