Physics > Geophysics
[Submitted on 13 May 2023]
Title:Physics reliable frugal uncertainty analysis for full waveform inversion
View PDFAbstract:Full waveform inversion (FWI) enables us to obtain high-resolution velocity models of the subsurface. However, estimating the associated uncertainties in the process is not trivial. Commonly, uncertainty estimation is performed within the Bayesian framework through sampling algorithms to estimate the posterior distribution and identify the associated uncertainty. Nevertheless, such an approach has to deal with complex posterior structures (e.g., multimodality), high-dimensional model parameters, and large-scale datasets, which lead to high computational demands and time-consuming procedures. As a result, uncertainty analysis is rarely performed, especially at the industrial scale, and thus, it drives practitioners away from utilizing it for decision-making. This work proposes a frugal approach to estimate uncertainty in FWI through the Stein Variational Gradient Descent (SVGD) algorithm by utilizing a relatively small number of velocity model particles. We warm-start the SVGD algorithm by perturbing the optimized velocity model obtained from a deterministic FWI procedure with random field-based perturbations. Such perturbations cover the scattering (i.e., high wavenumber) and the transmission (i.e., low wavenumber) components of FWI and, thus, represent the uncertainty of the FWI holistically. We demonstrate the proposed approach on the Marmousi model; we have learned that by utilizing a relatively small number of particles, the uncertainty map presents qualitatively reliable information that honours the physics of wave propagation at a reasonable cost, allowing for the potential for industrial-scale applications. Nevertheless, given that uncertainties are underestimated, we must be careful when incorporating them into downstream tasks of seismic-driven geological and reservoir modelling.
Submission history
From: Muhammad Izzatullah [view email][v1] Sat, 13 May 2023 14:06:01 UTC (5,627 KB)
Current browse context:
physics.geo-ph
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.