“Automatic velocity model building with complex salt: Can computers finally do an interpreter's job?”

Michell, Scott; Shen, Xuan; Brenders, Andrew; Dellinger, Joe; Ahmed, Imtiaz; Fu, Kang

doi:10.1190/segam2017-17778443.1

Cited by 36 publications

(6 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on our experience applying our new FWI workflow to many different data sets, we think the vision of the future described by Michell et al (2017) is likely correct: sparse OBN data with ultra-long offsets and a good low-frequency source may provide an economic yet adequate solution for automatic VMB using FWI in complex salt areas (Dellinger et al, 2016;Shen et al, 2018), with much less human interpretation than we have been practicing for the last two decades.…”

Section: Discussionmentioning

confidence: 98%

Correcting for salt misinterpretation with full-waveform inversion

Zhang

Mei

Lin

et al. 2018

SEG Technical Program Expanded Abstracts 2018

View full text Add to dashboard Cite

Using full-waveform inversion (FWI) to update velocity models that contain salt bodies with high velocity contrasts is challenging. It is even harder if erroneous salt geometry is part of the velocity model. Shen et al. (2017) showed a successful FWI application that corrected some misinterpretation of salt structures and resulted in improved subsalt images at the Atlantis field in the Gulf of Mexico. Their study stressed the importance of the low frequencies (usable down to 1.6 Hz), full azimuths, and long offsets of OBN data. Encouraged by the success at Atlantis, we revisited some aspects of FWI algorithms to minimize cycle-skipping and amplitude discrepancy issues that are common in the presence of salt and salt misinterpretation. Here we present the use of traveltime misfit measured in frequency-dependent time windows as the FWI cost function. It is devised to minimize the negative impact from the amplitude discrepancy and cycleskipping between the recorded data and modeled synthetic data. Furthermore, we use the crosscorrelation coefficient between the recorded data and synthetic data as a weight function in gradient computation to promote traveltime measurements of higher quality. We demonstrate the effect of our approach using a full-azimuth streamer data set in an area of complex shallow salt bodies in the Gulf of Mexico.

show abstract

Section: Discussionmentioning

confidence: 98%

Correcting for salt misinterpretation with full-waveform inversion

Zhang

Mei

Lin

et al. 2018

SEG Technical Program Expanded Abstracts 2018

View full text Add to dashboard Cite

show abstract

“…Current velocity model building strategies lean towards a more approximate salt interpretation and letting computer algorithms cope with the refinement. Michell et al (2017) and Shen et al (2018) presented some recent examples of the successful application of full-waveform inversion to simultaneously update sediments and salt in the Gulf of Mexico using OBN data. A notable example in the Brazilian Santos Basin, with NATS data, is the use of time-lag full waveform inversion (TLFWI) (Zhang et al, 2018b) followed by image-domain reflection full waveform inversion (IRFWI).…”

Section: Mitigating Interpretation Uncertainties With Fwimentioning

confidence: 99%

“…Progress in FWI have promoted direct shaping of the salt, for example Zhang et al, (2018) and Michell et al (2017). The updates given by FWI can also be used as a guide for new salt scenarios.…”

Section: Introductionmentioning

confidence: 99%

Status of salt interpretation for velocity model building in Brazilian basins

Marpeau¹,

Belz²

2019

Proceedings of the 16th International Congress of the Brazilian Geophysical Society&Expogef

View full text Add to dashboard Cite

show abstract

“…These attempts often require both geophysical processing skills and geological interpretation (Farmer et al, 1996;Jones, 2012;McCann et al, 2012). Shen et al (2017) and Michell et al (2017) applied FWI to successfully correct some salt structure misinterpretation based on the availability of low frequency and full azimuth of ocean-bottom node (OBN) data, which greatly improved subsalt delineation in the Gulf of Mexico (GOM). Zhang et al (2018) proposed a time-lag FWI (TLFWI) which uses a time-lag cost function (Luo and Schuster, 1991;Chavent et al, 1994) to reduce amplitude errors in the standard datamismatch cost function (Tarantola, 1984).…”

Section: Introductionmentioning

confidence: 99%

3D anisotropic full-waveform inversion for complex salt provinces

Rusmanugroho

Kalita³

et al. 2023

Front. Earth Sci.

View full text Add to dashboard Cite

Seismic imaging and inversion become extremely challenging when dealing with salt structures. Conventional state-of-the-art full-waveform inversion (FWI) fails to recover those features in areas where salt is present. A widely used solution in industry, however, involves substantial human interpretation. In this study, a regularized isotropic full-waveform inversion that penalizes the velocity drops in depth is introduced to recover the top parts of salt bodies. Then, an automatic salt flooding is applied to reconstruct deeper parts of the salt. Finally, in order to improve the accuracy of the inverted model, which is strongly affected by anisotropy, an anisotropic FWI is used to update the velocity model. We tested the approach on both synthetic and field datasets. Our FWI results revealed satisfying salt recovery as well as detailed velocity features in areas close to salt bodies.

show abstract

“Automatic velocity model building with complex salt: Can computers finally do an interpreter's job?”

Cited by 36 publications

References 11 publications

Correcting for salt misinterpretation with full-waveform inversion

Correcting for salt misinterpretation with full-waveform inversion

Status of salt interpretation for velocity model building in Brazilian basins

3D anisotropic full-waveform inversion for complex salt provinces

Contact Info

Product

Resources

About