Seismic Waveform Inversion Using The Ensemble Kalman Smoother

Author(s):  
M. Gineste ◽  
J. Eidsvik
Author(s):  
Michael Gineste ◽  
Jo Eidsvik

AbstractAn ensemble-based method for seismic inversion to estimate elastic attributes is considered, namely the iterative ensemble Kalman smoother. The main focus of this work is the challenge associated with ensemble-based inversion of seismic waveform data. The amount of seismic data is large and, depending on ensemble size, it cannot be processed in a single batch. Instead a solution strategy of partitioning the data recordings in time windows and processing these sequentially is suggested. This work demonstrates how this partitioning can be done adaptively, with a focus on reliable and efficient estimation. The adaptivity relies on an analysis of the update direction used in the iterative procedure, and an interpretation of contributions from prior and likelihood to this update. The idea is that these must balance; if the prior dominates, the estimation process is inefficient while the estimation is likely to overfit and diverge if data dominates. Two approaches to meet this balance are formulated and evaluated. One is based on an interpretation of eigenvalue distributions and how this enters and affects weighting of prior and likelihood contributions. The other is based on balancing the norm magnitude of prior and likelihood vector components in the update. Only the latter is found to sufficiently regularize the data window. Although no guarantees for avoiding ensemble divergence are provided in the paper, the results of the adaptive procedure indicate that robust estimation performance can be achieved for ensemble-based inversion of seismic waveform data.


2019 ◽  
Vol 100 (1) ◽  
pp. 313-327
Author(s):  
Dan Yu ◽  
Xinghui Huang ◽  
Zhengyuan Li

Abstract A catastrophic landslide struck the Xiaoba village in Fuquan, Guizhou, southwestern China at about 8:30 p.m. (Beijing Time, UTC + 8) on August 27, 2014. The landslide and induced impulse water waves destroyed two villages and killed 23 persons. By reprocessing seismic signals from a seismic network deployed in the surrounding area of the landslide, we recognized the event from low-frequency seismic signals and subsequently performed a long-period seismic waveform inversion to obtain its force–time history. The inversion results reveal that the maximum force for the landslide is 5 × 109 N, and the duration of the landslide is 38.4 s. The landslide reached its maximum velocity of 12.4 m/s at 13.2 s after its initiation, and the mass center plugged into the quarry at 24.2 s. Based on the inversion results, we estimated basal friction of the landslide. We found the friction coefficient rapidly reduces to a relatively steady-state value of ~ 0.4 at a steady-state distance of 35 m and subsequently reduces in a near-linear manner that satisfies the empirical formula $$ \mu = - 1.4d + 0.44 $$μ=-1.4d+0.44, where $$ d $$d is sliding distance in km. The reduction in friction revealed by the formula is compatible with the finding of previous studies for landslides of similar volume in landslide acceleration stage. However, our result does not make it possible for the friction coefficient to increase again in landslide deceleration stage that a velocity-dependent friction law would allow. The friction variation patterns can be used to constrain input parameters in numerical landslide simulation, which can predicate runout distance and deposit areas for massive landslides to carry out landslide hazard assessment.


Sign in / Sign up

Export Citation Format

Share Document