Bayesian Inverse Methods for Ice Sheet Model Calibration and Reconstruction
Calibrating ice sheet model SICOPOLIS using age layer radar data in order to project the future impact on sea level accurately.
Deep Learning Emulator for the dynamics of sea ice within the MITgcm model
Can we reduce the computational costs of dynamic sea ice models using Deep Learning?
Laplacian 2D Finite Difference (FD) solver on Stampede2 Supercomputer
Leveraged TACC’s Stampede2 Supercomputer to build a heat equation solver.
Deep Mountain Glaciers
Leveraging Deep Learning to emulate as well as invert non-linear, highly diffusive mountain glacier model.
Solving PDE-constrained inverse problems using FEniCS
Solved ill-conditioned inverse problems using the adjoint equations and special second order methods.
Machine Learning applications in geophysics
Explored utility of autoencoders, GANs, CNNs, ML algorithms across multiple applications in geophysics.
Uncertainty Quantification in Computational Models
Using Bayesian framework to quantify uncertainties in model parameters.