Towards a More Robust Joint Migration Inversion

Abolhassani, S.; Verschuur, Eric

doi:10.3997/2214-4609.202210383

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There have been multiple attempts so far to approximate a proper but cheaper preconditioner (e.g., Jun et al, 2015;Lu et al, 2018). Recently, Abolhassani and Verschuur (2022a) proposed a cheap, depth-based, data-domain preconditioner in the context of iterative least-squares one-way wave-equation migration (LS-WEM) for image reconstruction. This expanded abstract will build on their proposal for a cheap, depth-based preconditioner for high-resolution background velocity reconstruction with ORWI.…”

Section: Introductionmentioning

confidence: 99%

High-Resolution One-Way Reflection Waveform Inversion

Abolhassani

Verschuur

2023

84th EAGE Annual Conference &Amp; Exhibition

View full text Add to dashboard Cite

Reflection waveform inversion (RWI) is a method that relies on primary pure reflection data to recover the subsurface background velocity based on the associated evolving seismic images. Background velocity updates estimated by conventional RWI are nonoptimal, which is partly attributed to lowresolution tomographic wavepaths and migration isochrones. Preconditioning RWI sensitivity kernels using Hessian information solves this problem but is not practical for a large number of model parameters. One-way reflection waveform inversion (ORWI) is a reflection waveform tomography technique in which the forward modeling scheme operates in one direction (downward and then upward) via virtual parallel data levels in the medium. The ORWI framework allows us to break down the Hessian matrix into smaller operators, which makes the preconditioning operation more efficient and less computationally expensive. This extended abstract turns conventional ORWI into a high-resolution but computationally feasible ORWI (Gauss-Newton ORWI) to improve the nonoptimal background velocity updates.

show abstract

Section: Introductionmentioning

confidence: 99%