TY - GEN
T1 - Bayesian uncertainty estimation for full waveform inversion: A numerical study
AU - Izzatullah, Muhammad
AU - van Leeuwen, Tristan
AU - Peter, Daniel
N1 - KAUST Repository Item: Exported on 2020-10-01
Acknowledgements: This research work was performed under supervision of Tristan van Leeuwen during the first author's research visit to Mathematical Institute of Utrecht University, Netherlands from 20th January to 5th February 2019. The first author would like to thank Tristan van Leeuwen for his guidance throughout the visit. The research visit and the work reported here was supported by funding from King Abdullah University of Science and Technology (KAUST). We would like to thank Nick Luiken, Ajinkya Kadu, and Sarah Gaaf from Mathematical Institute of Utrecht University, Netherlands and members of the Seismic Modeling and Inversion (SMI) group at KAUST especially Qiancheng Liu, Armando Espindola and Diaz Urozayev for constructive discussions.
PY - 2019/8/10
Y1 - 2019/8/10
N2 - Full waveform inversion enables us to obtain high-resolution subsurface images. However, estimating the associated uncertainties is not trivial. Hessian-based method gives us an opportunity to assess the uncertainties around a given estimate based on the inverse of the Hessian, evaluated at that estimate. In this work we study various algorithms for extracting information from this inverse Hessian based on a low-rank approximation. In particular, we compare the Lanczos method to the randomized singular value decomposition. We demonstrate that the low-rank approximation may lead to a biased conclusion.
AB - Full waveform inversion enables us to obtain high-resolution subsurface images. However, estimating the associated uncertainties is not trivial. Hessian-based method gives us an opportunity to assess the uncertainties around a given estimate based on the inverse of the Hessian, evaluated at that estimate. In this work we study various algorithms for extracting information from this inverse Hessian based on a low-rank approximation. In particular, we compare the Lanczos method to the randomized singular value decomposition. We demonstrate that the low-rank approximation may lead to a biased conclusion.
UR - http://hdl.handle.net/10754/661895
UR - https://library.seg.org/doi/10.1190/segam2019-3216008.1
UR - http://www.scopus.com/inward/record.url?scp=85079488113&partnerID=8YFLogxK
U2 - 10.1190/segam2019-3216008.1
DO - 10.1190/segam2019-3216008.1
M3 - Conference contribution
SP - 1685
EP - 1689
BT - SEG Technical Program Expanded Abstracts 2019
PB - Society of Exploration Geophysicists
ER -