ABSTRACT

Speeding-up convergence rates and reducing the computational burden of Full Waveform Inversion (FWI) is increasingly important as we move toward large-scale 3D multi-parameter inversion. To this end, second-order optimization algorithms like L-BFGS or the truncated Newton method allow a much faster convergence rate at minimal computational costs. In the same fashion, stochastic source subsampling approaches have been shown to reduce the computational cost of FWI. In this study, we propose to combine these two strategies and present how the L-BFGS algorithm can be used along with the stochastic source subsampling strategy, or what we call the stochastic L-BFGS algorithm.

Presentation Date: Tuesday, September 26, 2017

Start Time: 2:40 PM

Location: Exhibit Hall C, E-P Station 3

Presentation Type: EPOSTER

This content is only available via PDF.
You can access this article if you purchase or spend a download.