I am primarily interested in developing techniques for seismogram synthesis and imaging, and estimation of rock properties from geophysical data. During my graduate studies I worked on asymptotic methods for seismogram synthesis. You can essentially choose one of the two methods for seismogram synthesis: Exact (or more exact) methods such as the reflectivity methods, which are limited to the type of earth model you can use or Asymptotic methods (such as Ray theory and its variants) where you can use fairly realistic earth models. With graduate student Jonas De Basabe, I worked on error analysis of the spectral element and DG finite element method. Prof. Yang Liu from China University of Petroleum and I are working on new strategies for FD modeling in elastic and anisotropic media.
Seismic Migration or Imaging is an art – they are based on wave propagation models. One with a good background in theory can contribute well in this field. Much of the ideas are intuitive – they work amazingly well. The potential of the plane wave methods for imaging has not been explored well. For example it is trivial to extend plane wave theory to include the effects of anisotropy and multi-component data. Much of my current research is focussed on this aspect.
Geophysical inverse problems are an active area of research. Since joining UTIG, I worked on several geophysical parameter estimation problems. We were able to demonstrate that global optimization methods such as simulated annealing (SA) and genetic algorithm (GA) can be used effectively to several geophysical applications such as the seismic waveform inversion problem. I also showed how one could apply Gibbs’ sampling and Metropolis rule to characterize uncertainties and the posterior probability density function (PPD) in model space. We are actively pursuing this area of research with application to seismic reflection data from gas hydrates, global seismology, seismic tomography, and anisotropic parameter estimation. We are developing efficient full waveform inversion algorithms for laterally varying media.
I develop parallel algorithms for seismic processing, and model earthquake seismograms for large scale earth structures. Jay Pulliam and I are studying seismograms from deep focus earthquakes to constrain anisotrpoic structures of D”.
Geophysics is a lot of fun. It is an exciting field to be in – it is truly interdisciplinary. We borrow ideas and techniques developed in Physics, Math, Statistics, Computer Science, Electrical Engg and Biology – to name a few, and apply them to several geophysical problems. Of course the main objective is to solve geologic problems. I would consider it a waste of time to develop a technique that does not work on real data.
Publications – 2020
Multifrequency beam-based migration in inhomogeneous media using windowed Fourier transform frames
R. Tuvi, Z. Zhao, M. K. Sen, Geophysical Journal International, 223 (2), 1086-1099
A gradient based Markov chain Monte Carlo method for full waveform inversion and uncertainty analysis
Z. Zhao, M., K. Sen, Geophysics, 86 (1), 1-79
Probabilistic joint-inversion of marine CSEM and seismic traveltime data using VFSA and generalized fuzzy clustering
P. K. Mishra, A. Arnulf, M. K. Sen, B. Denel, Y. Sun, P. Williamson, SEG Technical Program Expanded Abstracts 2020, 545-549
Full-waveform inversion using machine learning optimization techniques
R. de Bragança, J. Vamaraju, M. K. Sen, SEG Technical Program Expanded Abstracts 2020, 865-869
Two-step velocity inversion using trans-dimensional tomography and elastic FWI
R. Biswas, A. F. Arnulf, M. K. Sen, D. Datta, Z. Zhao, P. K. Mishra, P. Jaysaval, SEG Technical Program Expanded Abstracts 2020, 3628-3633
Meshfree seismic modeling using radial basis finite-difference with adaptive stencil size
X. Liu, P. K. Mishra, M. K. Sen, L. Ling, SEG Technical Program Expanded Abstracts 2020, 2673-2677
A fast image domain least squares migration method with local data target approach
R. Tuvi, Z. Zhao, M. K. Sen, SEG Technical Program Expanded Abstracts 2020, 2893-2897
Anisotropy correction and fluid substitution for deviated wells
S. Phan, M. K. Sen, SEG Technical Program Expanded Abstracts 2020, 190-194
Optimal mini-batch shot selection for target-oriented least-squares reverse time migration
J. Vamaraju, R. Tuvi, Z. Zhao, M. K. Sen, SEG Technical Program Expanded Abstracts 2020, 3073-3077
Wave propagation with physics informed neural networks
D. Voytan, M. K. Sen, SEG Technical Program Expanded Abstracts 2020, 3477-3481
Facies and reservoir properties estimation by a transdimensional seismic inversion
A. Dhara, M. K. Sen, D. Yang, J. Schmedes, P. Routh, R. Sain, SEG Technical Program Expanded Abstracts 2020, 255-259
A hybrid Galerkin finite element method for seismic wave propagation in fractured media
J. Vamaraju, M. K. Sen, J. De Basabe, M. Wheeler, Geophysical Journal International, 221 (2), 857-878
Adaptive radial basis function generated finite-difference (RBF-FD) on non-uniform nodes using p-refinement
P. K. Mishra, L. Ling, X. Liu, M. K. Sen, arXiv preprint
Frequency-dependent AVO analysis using the scattering response of a layered reservoir
D. Kumar, Z. Zhao, D. J. Foster, D. Dralus, M. K. Sen, Geophysics, 85 (2), N1-N16
Frequency-dependent AVO analysis
D. J. Foster, Z. Zhao, D. Kumar, D. Dralus, M. K. Sen, The Leading Edge, 39 (2), 84-91
Effective Medium Modeling of CO2-Sequestered Carbonate Reservoir
R. Ghosh, M. K. Sen, Petro-physics and Rock Physics of Carbonate Reservoirs, 145-160
Inversion of downhole electrical measurements for proppant mapping using very fast simulated annealing
P. Zhang, J. Shiriyev, M. K. Sen, M. M. Sharma, Geophysics, 85 (1), D13-D22
Publications – 2019
Understanding the encoder-decoder network for solving an unsupervised seismic inversion problem
D. Barman, R. Biswas, M. K. Sen, AGUFM 2019, S51B-08
Mini-batch Least-Squares Reverse Time Migration In A Deep Learning Framework
J. Vamaraju, J. Vila, M. Araya-Polo, D. Datta, M. Sidahmed, M. K. Sen, AGUFM 2019, S51B-03
Full waveform inversion with a gradient based Markov chain Monte Carlo method and uncertainties quantification, application on Axial seamount
M. K. Sen, Z. Zhao, A. F. Arnulf, AGUFM 2019, S44B-05
Accelerating Least Squares Imaging Using Deep Learning Techniques
J. Vamaraju, J. Vila, M. Araya-Polo, D. Datta, M. Sidahmed, M. K. Sen, arXiv preprint
Estimating normal moveout velocity using the recurrent neural network
R. Biswas, A. Vassiliou, R. Stromberg, M. K. Sen, Interpretation, 7 (4), T819-T827
Frequency‐domain double‐plane‐wave least‐squares reverse time migration
Z. Zhao, M. K. Sen, Geophysical Prospecting, 67 (8), 2061-2084
Fast 2D full-waveform modeling and inversion using the Schur complement approach
D. Datta, P. Jaysaval, M. K. Sen, A. F. Arnulf, Geophysics, 84 (5), R783-R792
Full-waveform inversion of salt models using shape optimization and simulated annealing
D. Datta, M. K. Sen, F. Liu, S. Morton, Geophysics, 84 (5), R793-R804
2.5D controlled-source electromagnetic inversion using very fast simulated annealing algorithm
P. Jaysaval, D. Datta, M. K. Sen, A. F. Arnulf, B. Denel, P. Williamson, SEG Technical Program Expanded Abstracts 2019, 1060-1064
A multi-scale full waveform inversion method-staging wavenumber components and layer-stripping
Z. Zhao, M. K. Sen, SEG Technical Program Expanded Abstracts 2019, 1470-1474
Anisotropic moveout correction using a Hough transform neural network
J. Vamaraju, M. K. Sen, SEG Technical Program Expanded Abstracts 2019, 2383-2387
Density inversion from seismic using a trans-dimensional approach: A field dataset example
R. Biswas, D. Kumar, M. K. Sen, A. Paul, K. Packer, SEG Technical Program Expanded Abstracts 2019, 539-543
Deep learning with cross-shape deep Boltzmann machine for pre-stack inversion problem
S. Phan, M. K. Sen, SEG Technical Program Expanded Abstracts 2019, 2293-2297
A gradient based MCMC method for FWI and uncertainty analysis
Z. Zhao, M. K. Sen, SEG Technical Program Expanded Abstracts 2019, 1465-1469
A phase-space beam summation imaging in inhomogeneous medium
R. Tuvi, Z. Zhao, M. K. Sen, SEG Technical Program Expanded Abstracts 2019, 4356-4360
Pre-stack target-oriented least squares RTM in image domain using conjugate Hopfield networks and the Boltzmann machine
J. Vamaraju, Z. Zhao, M. K. Sen, SEG Technical Program Expanded Abstracts 2019, 2308-2312
Pre-stack inversion using a physics-guided convolutional neural network
R. Biswas, M. K. Sen, V. Das, T. Mukerji, SEG Technical Program Expanded Abstracts 2019, 4967-4971
Prestack and poststack inversion using a physics-guided convolutional neural network
R. Biswas, M. K. Sen, V. Das, T. Mukerji, Interpretation, 7 (3), SE161-SE174
A Boltzmann machine for high-resolution prestack seismic inversion
S. D. T. Phan, M. K. Sen, Interpretation, 7 (3), SE215-SE224
Unsupervised physics-based neural networks for seismic migration
J. Vamaraju, M. K. Sen, Interpretation, 7 (3), SE189-SE200
Finite Difference Based Wave Simulation in Fractured Porous Rocks
J. Vamaraju, M .K. Sen, arXiv preprint
Least-squares path-summation diffraction imaging using sparsity constraints
D. Merzlikin, S. Fomel, M. K. Sen, Geophysics, 84 (3), S187-S200
A stabilized radial basis-finite difference (RBF-FD) method with hybrid kernels
P .K. Mishra, G .E. Fasshauer, M .K. Sen, L. Ling, Computers & Mathematics with Applications, 77 (9), 2354-2368
Lithospheric removal beneath the eastern flank of the Rio Grande Rift from receiver function velocity analysis
M. Agrawal, J. Pulliam, M. K. Sen, S. P. Grand, Geochemistry, Geophysics, Geosystems, 20 (2), 974-991
Estimating depths to subsurface discontinuities using receiver function velocity analysis
M. Agrawal, J. Pulliam, M. K. Sen, Geophysical Research Abstracts, 21