Conditional Image Prior for Uncertainty Quantification in Full Waveform Inversion

TL;DR

This paper introduces a conditional CNN-based image prior for FWI to quantify model uncertainties, enabling efficient sampling from the posterior distribution and improving uncertainty assessment in high-dimensional seismic inversion.

Contribution

It proposes a novel conditional CNN prior that learns from Gaussian Random Field perturbations and adapts during FWI to generate posterior samples for uncertainty quantification.

Findings

Effective in modeling posterior distributions in FWI

Able to generate multiple plausible models

Demonstrated on Marmousi and field data

Abstract

Full Waveform Inversion (FWI) is a technique employed to attain a high resolution subsurface velocity model. However, FWI results are effected by the limited illumination of the model domain and the quality of that illumination, which is related to the quality of the data. Additionally, the high computational cost of FWI, compounded by the high dimensional nature of the model space, complicates the evaluation of model uncertainties. Recent work on applying neural networks to represent the velocity model for FWI demonstrated the network's ability to capture the salient features of the velocity model. The question we ask here is how reliable are these features in representing the observed data contribution within the model space (the posterior distribution). To address this question, we propose leveraging a conditional Convolutional Neural Network (CNN) as image prior to quantify the neural network uncertainties. Specifically, we add to the deep image prior concept a conditional channel, enabling the generation of various models corresponding to the specified condition. We initially train the conditional CNN to learn (store) samples from the prior distribution given by Gaussian Random Fields (GRF) based perturbations of the current velocity model. Subsequently, we use FWI to update the CNN model representation of the priors so that it can generate samples from the posterior distribution. These samples can be used to measure the approximate mean and standard deviation of the posterior distribution, as well as draw samples representing the posterior distribution. We demonstrate the effectiveness of the proposed approach on the Marmousi model and in a field data application.