Variational Bayesian experimental design for geophysical applications: seismic source location, amplitude versus offset inversion, and estimating CO2 saturations in a subsurface reservoir

Abstract

This paper introduces variational design methods that are novel to Geophysics, and discusses their benefits and limitations in the context of geophysical applications and more established design methods. Variational methods rely on functional approximations to probability distributions and model-data relationships. They can be used to design experiments that best resolve either all model parameters, or the answer to specific questions about the system to be interrogated. The methods are tested in three schematic geophysical applications: (i) estimating a source location given arrival times at sensor locations, and (ii) estimating the contrast in seismic wavefield velocity across a stratal interface given measurements of the amplitudes of seismic wavefield reflections from that interface, and (iii) designing a survey to best constrain CO2 saturation in a subsurface storage scenario. Variational methods allow the value of an experiment to be calculated and optimised simultaneously, which results in substantial savings in computational cost. In the context of designing a survey to best constrain CO2 in a subsurface storage scenario, we show that optimal designs may change substantially depending on the questions of interest. Overall, this work indicates that optimal design methods should be used more widely in Geophysics, as they are in other scientifically advanced fields.

arXiv commentSubmitted to Geophysical Journal International in June 2023

References in corpus (14)

• Adam: A Method for Stochastic Optimization
• Array Programming with NumPy
• NICE: Non-linear Independent Components Estimation
• Variational Inference with Normalizing Flows
• Simulation-based optimal Bayesian experimental design for nonlinear systems
• Gradient-based stochastic optimization methods in Bayesian experimental design
• Fast Bayesian experimental design: Laplace-based importance sampling for the expected information gain
• Goal-Oriented Optimal Design of Experiments for Large-Scale Bayesian Linear Inverse Problems
• Fast Bayesian Optimal Experimental Design for Seismic Source Inversion
• A fast and scalable computational framework for large-scale and high-dimensional Bayesian optimal experimental design
• A layered multiple importance sampling scheme for focused optimal Bayesian experimental design
• A Unified Stochastic Gradient Approach to Designing Bayesian-Optimal Experiments
• Tight Mutual Information Estimation With Contrastive Fenchel-Legendre Optimization
• Optimal Bayesian design for model discrimination via classification

Cited by in corpus (6)

• Variational Prior Replacement in Bayesian Inference and Inversion
• Variational Sequential Optimal Experimental Design using Reinforcement Learning
• The Roles of Low-Noise Stations, Arrays and Ocean-Bottom Seismometers in Monitoring UK Offshore Seismicity associated with Subsurface Storage of Carbon Dioxide
• A Multi-fidelity Estimator of the Expected Information Gain for Bayesian Optimal Experimental Design
• Single and Multi-Objective Optimization of Distributed Acoustic Sensing Cable Layouts for Geophysical Applications
• A decade of airborne electromagnetic surveying Lake Menindee (Australia) under varying water levels