A robust first-arrival picking workflow using convolutional and recurrent neural networks

Geophysics ◽  
2020 ◽  
Vol 85 (5) ◽  
pp. U109-U119
Author(s):  
Pengyu Yuan ◽  
Shirui Wang ◽  
Wenyi Hu ◽  
Xuqing Wu ◽  
Jiefu Chen ◽  
...  
Keyword(s):  
Deep Learning ◽  
Model Building ◽  
Signal To Noise Ratio ◽  
Synthetic Data ◽  
Velocity Model ◽  
Surface Velocity ◽  
Data Sets ◽  
Near Surface ◽  
Segmentation Boundary ◽  
First Arrival

A deep-learning-based workflow is proposed in this paper to solve the first-arrival picking problem for near-surface velocity model building. Traditional methods, such as the short-term average/long-term average method, perform poorly when the signal-to-noise ratio is low or near-surface geologic structures are complex. This challenging task is formulated as a segmentation problem accompanied by a novel postprocessing approach to identify pickings along the segmentation boundary. The workflow includes three parts: a deep U-net for segmentation, a recurrent neural network (RNN) for picking, and a weight adaptation approach to be generalized for new data sets. In particular, we have evaluated the importance of selecting a proper loss function for training the network. Instead of taking an end-to-end approach to solve the picking problem, we emphasize the performance gain obtained by using an RNN to optimize the picking. Finally, we adopt a simple transfer learning scheme and test its robustness via a weight adaptation approach to maintain the picking performance on new data sets. Our tests on synthetic data sets reveal the advantage of our workflow compared with existing deep-learning methods that focus only on segmentation performance. Our tests on field data sets illustrate that a good postprocessing picking step is essential for correcting the segmentation errors and that the overall workflow is efficient in minimizing human interventions for the first-arrival picking task.

A layer-stripping method for 3D near-surface velocity model building using seismic first-arrival times

2015 ◽  
Vol 26 (4) ◽  
pp. 502-507 ◽  
Author(s):  
Taikun Shi ◽  
Jianzhong Zhang ◽  
Zhonglai Huang ◽  
Changkun Jin
Keyword(s):  
Model Building ◽  
Velocity Model ◽  
Surface Velocity ◽  
Near Surface ◽  
Arrival Times ◽  
Velocity Model Building ◽  
Stripping Method ◽  
First Arrival

Robust refraction statics solution and near-surface velocity model building using feedback from reflection data

Geophysics ◽  
2018 ◽  
Vol 83 (6) ◽  
pp. U63-U77
Author(s):  
Bernard K. Law ◽  
Daniel Trad
Keyword(s):  
Model Building ◽  
Waveform Inversion ◽  
Velocity Model ◽  
Surface Velocity ◽  
Near Surface ◽  
Arrival Times ◽  
Reflection Data ◽  
Velocity Model Building ◽  
First Arrival ◽  
Refraction Data

An accurate near-surface velocity model is critical for weathering statics correction and initial model building for depth migration and full-waveform inversion. However, near-surface models from refraction inversion often suffer from errors in refraction data, insufficient sampling, and over-simplified assumptions used in refraction algorithms. Errors in refraction data can be caused by picking errors resulting from surface noise, attenuation, and dispersion of the first-arrival energy with offset. These errors are partially compensated later in the data flow by reflection residual statics. Therefore, surface-consistent residual statics contain information that can be used to improve the near-surface velocity model. We have developed a new dataflow to automatically include median and long-wavelength components of surface-consistent reflection residual statics. This technique can work with any model-based refraction solution, including grid-based tomography methods and layer-based methods. We modify the cost function of the refraction inversion by adding model and data weights computed from the smoothed surface-consistent residual statics. By using an iterative inversion, these weights allow us to update the near-surface velocity model and to reject first-arrival picks that do not fit the updated model. In this nonlinear optimization workflow, the refraction model is derived from maximizing the coherence of the reflection energy and minimizing the misfit between model arrival times and the recorded first-arrival times. This approach can alleviate inherent limitations in shallow refraction data by using coherent reflection data.

Tomographic velocity model building of the near surface with velocity-inversion interfaces: A test using the Yilmaz model

Geophysics ◽  
2010 ◽  
Vol 75 (6) ◽  
pp. U39-U47 ◽  
Author(s):  
Hui Liu ◽  
Hua-wei Zhou ◽  
Wenge Liu ◽  
Peiming Li ◽  
Zhihui Zou
Keyword(s):  
Model Building ◽  
Velocity Model ◽  
Surface Velocity ◽  
Low Velocity ◽  
Near Surface ◽  
Velocity Inversion ◽  
Velocity Models ◽  
Velocity Model Building ◽  
First Arrival ◽  
The Impact

First-arrival traveltime tomography is a popular approach to building the near-surface velocity models for oil and gas exploration, mining, geoengineering, and environmental studies. However, the presence of velocity-inversion interfaces (VIIs), across which the overlying velocity is higher than the underlying velocity, might corrupt the tomographic solutions. This is because most first-arrival raypaths will not traverse along any VII, such as the top of a low-velocity zone. We have examined the impact of VIIs on first-arrival tomographic velocity model building of the near surface using a synthetic near-surface velocity model. This examination confirms the severe impact of VIIs on first-arrival tomography. When the source-to-receiver offset is greater than the lateral extent of the VIIs, good near-surface velocity models can still be established using a multiscale deformable-layer tomography (DLT), which uses a layer-based model parameterization and a multiscale scheme as regularization. Compared with the results from a commercial grid-based tomography, the DLT delivers much better near-surface statics solutions and less error in the images of deep reflectors.

Integrated turning-ray and reflection tomography for velocity model building in foothill areas

2018 ◽  
Vol 6 (4) ◽  
pp. SM63-SM70 ◽  
Author(s):  
Tian Jun ◽  
Peng Gengxin ◽  
Junru Jiao ◽  
Grace (Yan) Yan ◽  
Xianhuai Zhu
Keyword(s):  
Model Building ◽  
Synthetic Data ◽  
Velocity Model ◽  
Surface Velocity ◽  
Seismic Exploration ◽  
Surface Model ◽  
Near Surface ◽  
Velocity Model Building ◽  
Tomographic Method ◽  
Reflection Tomography

A special challenge for land seismic exploration is estimating velocities, in part due to complex near-surface structures, and in some instances because of rugose topography over foothills. We have developed an integrated turning-ray and reflection-tomographic method to face this challenge. First, turning-ray tomography is performed to derive a near-surface velocity-depth model. Then, we combine the near-surface model with the initial-subsurface model. Taking the combined model as starting model, we go through a reflection tomographic process to build the model for imaging. During reflection tomography, the near-surface model and subsurface models are jointly updated. Our method has been successfully applied to a 2D complex synthetic data example and a 3D field data example. The results demonstrate that our method derives a very decent model even when there is no reflection information available in a few hundred meters underneath the surface. Joint tomography can lead to geologic plausible models and produce subsurface images with high fidelity.

Full waveform inversion: early start of model building process to resolve the complex near surface model in the Exmouth sub-basin, North West Shelf Australia

2018 ◽  
Vol 58 (2) ◽  
pp. 884
Author(s):  
Lianping Zhang ◽  
Haryo Trihutomo ◽  
Yuelian Gong ◽  
Bee Jik Lim ◽  
Alexander Karvelas
Keyword(s):  
Model Building ◽  
Waveform Inversion ◽  
Velocity Model ◽  
Surface Velocity ◽  
Image Point ◽  
Surface Model ◽  
Full Waveform Inversion ◽  
Near Surface ◽  
North West ◽  
Full Waveform

The Schlumberger Multiclient Exmouth 3D survey was acquired over the Exmouth sub-basin, North West Shelf Australia and covers 12 600 km2. One of the primary objectives of this survey was to produce a wide coverage of high quality imaging with advanced processing technology within an agreed turnaround time. The complexity of the overburden was one of the imaging challenges that impacted the structuration and image quality at the reservoir level. Unlike traditional full-waveform inversion (FWI) workflow, here, FWI was introduced early in the workflow in parallel with acquisition and preprocessing to produce a reliable near surface velocity model from a smooth starting model. FWI derived an accurate and detailed near surface model, which subsequently benefitted the common image point (CIP) tomography model updates through to the deeper intervals. The objective was to complete the FWI model update for the overburden concurrently with the demultiple stages hence reflection time CIP tomography could start with a reasonably good velocity model upon completion of the demultiple process.

Model–parameterization scaling for improving accuracy of seismic tomography in reconstructing near–surface velocity model

2020 ◽  
Author(s):  
Gleb S. Chernyshov ◽  
◽  
Anton A. Duchkov ◽  
Aleksander A. Nikitin ◽  
Ivan Yu. Kulakov ◽  
...  
Keyword(s):  
Model Building ◽  
Regularization Parameter ◽  
Velocity Model ◽  
Surface Velocity ◽  
Surface Model ◽  
Tomographic Inversion ◽  
Near Surface ◽  
Model Parameterization ◽  
Regularization Parameters ◽  
Improving Accuracy

The problem of tomographic inversion is non–unique and requires regularization to solve it in a stable manner. It is highly non–trivial to choose between various regularization approaches or tune the regularization parameters themselves. We study the influence of one particular regularization parameter on the resolution and accuracy the tomographic inversion for the near–surface model building. We propose another regularization parameter, which allows to increase the accuracy of model building.

Sign in / Sign up

Export Citation Format

Share Document