A reality check on full-wave inversion applied to land seismic data for near-surface modeling

We evaluate the performance of traveltime tomography and full-wave inversion (FWI) for near-surface modeling using the data from a shallow seismic field experiment. Eight boreholes up to 20-m depth have been drilled along the seismic line traverse to verify the accuracy of the P-wave velocity-depth model estimated by seismic inversion. The velocity-depth model of the soil column estimated by traveltime tomography is in good agreement with the borehole data. We used the traveltime tomography model as an initial model and performed FWI. Full-wave acoustic and elastic inversions, however, have failed to converge to a velocity-depth model that desirably should be a high-resolution version of the model estimated by traveltime tomography. Moreover, there are significant discrepancies between the estimated models and the borehole data. It is understandable why full-wave acoustic inversion would fail — land seismic data inherently are elastic wavefields. The question is: Why does full-wave elastic inversion also fail? The strategy to prevent full-wave elastic inversion of vertical-component geophone data trapped in a local minimum that results in a physically implausible near-surface model may be cascaded inversion. Specifically, we perform traveltime tomography to estimate a P-wave velocity-depth model for the near-surface and Rayleigh-wave inversion to estimate an S-wave velocity-depth model for the near-surface, then use the resulting pairs of models as the initial models for the subsequent full-wave elastic inversion. Nonetheless, as demonstrated by the field data example here, the elastic-wave inversion yields a near-surface solution that still is not in agreement with the borehole data. Here, we investigate the limitations of FWI applied to land seismic data for near-surface modeling.

Remark on Algorithm 992: An OpenGL- and C++-based Function Library for Curve and Surface Modeling in a Large Class of Extended Chebyshev Spaces

We provide a number of corrections to the software component that accompanied this Algorithm submission [3]. An updated version of the code is available from the ACM Collected Algorithms site [1].

Full Waveform Inversion for the Near Surface Characterization, Onshore UAE, Case Study

Full-waveform inversion (FWI) has become a well-established method for obtaining a detailed earth model suitable for improved imaging, near-surface characterization and pore-pressure prediction. FWI for onshore data has always been challenging and has seen limited application (Vigh et al, 2018). It requires a dedicated data processing approach related to the lower signal-to-noise ratio, accounting for variable topography and complex near-surface related effects. During the past few years, ADNOC has been acquiring and processing one of the world's largest combined 3D onshore and offshore seismic surveys in the Emirate of Abu Dhabi. The modern acquisition parameters that were implemented enabled the acquisition of broadband onshore seismic data rich in low frequencies that could benefit the initial stages of the FWI workflow. Sand dunes and sabkha layers at the surface, and high-velocity carbonate and dolomite layers in the subsurface pose a significant challenge for near-surface modeling in the UAE. The purpose of this work is to evaluate FWI application onshore UAE for near-surface characterization. We will compare the FWI results with conventional approaches for the near-surface model building that has been used routinely on land datasets in UAE, such as data-driven image-based statics (DIBS, Zarubov et al, 2019). One of the main challenges is data preconditioning, as onshore seismic data typically exhibits high levels of noise. It is imperative to denoise gathers sufficiently prior to the FWI process. A well sonic velocity function with large smoothing was used to build the starting velocity model for FWI. The process aims to minimize the least-squared difference between predicted and observed seismic responses by means of updating the model on which the prediction is based. As the predicted and seismic responses are functions of model parameters as well as source signature, a good estimate of the source wavelet is important for update and convergence in FWI. During this FWI work, source wavelet inversion was done as a separate step and used in subsequent FWI passes. FWI inversion started with adjustive FWI (Kun et al, 2015) on lower frequencies, moving to higher frequencies where both adjustive and least square objective functions were used. We will further show assessment of the anisotropy, initial conditions, usage of geological constraints, and comparisons to the conventional solutions. A comparison of results shows that FWI has successfully added velocity details to the near-surface model that follow the geological trend and conforms to well information while producing a plausible static solution. We have demonstrated the application of FWI onshore UAE for near-surface modeling. Although turnaround time (TAT) has increased compared to the conventional approach, the learning that was gained during this trial will decrease TAT for the future FWI work.