Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
SUMMARY Using the Hessian matrix to scale the gradients in full-waveform inversion can reduce the artifacts caused by the geometrical spreading and defocusing effects caused by incomplete data coverage. However, it is computationally expensive to calculate the Hessian matrix. We develop a new wave-energy-based precondition method for full-waveform inversion to reduce the artifacts in the gradients caused by the geometrical spreading and defocusing effects. This method scales the gradients using the square root of the wave energy of the forward propagated wavefields from sources and that of backpropagated synthetic wavefields from receivers. We use a complex model to validate the improvements of our new wave-energy-based precondition method for full-waveform inversion. Our results show that our new method provides a faster convergence rate and converges to a more accurate result than the existing precondition methods, particularly for the deep region of the model.
- Research Report > New Finding (0.55)
- Overview > Innovation (0.35)
SUMMARY We present a differential structural constraint for joint inversion of regional scale seismic traveltimes and magne-totelluric (MT) data, resulting in a simple and computationally efficient inversion algorithm. We discuss the algorithm and test results for a synthetic data, featuring high contrast blocky basement and salt dome in sediments.
- Geophysics > Electromagnetic Surveying (1.00)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling > Seismic Inversion (0.51)
Summary For a robust elastic waveform inversion algorithm, we propose incorporating a denoise function into gradients in the l1-norm waveform inversion. The denoise function is designed by the ratio of modeled data to field data summed over shots and receivers at each frequency, based on the fact that while field data are noisy, modeled data are noise-free. As a result, the denoise function is inversely proportional to the degree of noises and acts like filters. Using the denoise function, we can keep the noise-contaminated gradients from affecting model parameter updates. The denoise function is applied to synthetic data with three types of noises for the modified version of Marmousi-2 model: discontinuous monochromatic random noises, general random noises and outliers. Numerical examples show that the denoise function effectively filters out the noise-contaminated gradients during the inversion process and thus yields better inversion results than the conventional l1-norm waveform inversion.
Summary The GRACE (Gravity Recovery and Climate Experiment) satellite data allow for project scale exploration for mineral deposits. The data provide good regional coverage in areas where no ground gravity data exist and can also be used to augment national databases. I used the data in two case studies. The first project was for potash exploration in the Bada Basin in eastern Eritrea. The GRACE gravity data were essential for establishing the regional gradient for modeling isolated ground gravity profiles across the basin. The second project was a gold exploration target in north central Nevada, USA. The GRACE satellite data showed a gravity high in the basin that was not indicated in the US National Geodetic Survey database. Detailed gravity surveying confirmed the GRACE gravity high was real and changed the geologic model for the project and is directing current exploration efforts.
- Energy > Oil & Gas > Upstream (1.00)
- Materials > Metals & Mining (0.91)
- North America > United States > Wyoming > Great Basin (0.99)
- North America > United States > Utah > Great Basin (0.99)
- North America > United States > Oregon > Great Basin (0.99)
- (3 more...)
Summary Full waveform inversion has been successful in building high resolution velocity models for shallow layers. To be successful, it requires refracted waves or low frequencies in the reflection/refraction data. We revisit full waveform inversion theory in hopes of relaxing the dependence on low frequency reflections. We implement an approach allowing the updating of long wavelength components of the velocity model affecting the reflected arrivals even with absence of low frequency in the input data. Our tactic is based on a non-linear iterative relaxation approach where short and long wavelength components of the velocity model are updated alternatively. We study theoretically the associated Fréchet derivatives and gradients and discussed how and why such a strategy improves the resolution that we can expect from full waveform inversion. The kernel of our approach is very similar to the algorithm of migration based travel time tomography proposed by Chavent et al. (1994). Finally we present a preliminary 2D application to a 2D Gulf of Mexico conventional streamer dataset.
- North America > United States (0.37)
- North America > Mexico (0.25)
SUMMARY Full waveform inversion (FWI) based upon least-squares is commonly solved through local optimization methods utilizing conjugate gradients or steepest descent. However, gradient optimization methods are generally considered slower and can have difficulties scaling the reconstructed model parameters in comparison to methods that require second-order information of the objective function, such as the Newton method. In this paper, we present an inexact full Newton optimization method for the full waveform inversion algorithm in the frequency domain which utilizes simultaneous sources based upon the phase encoding technique. Tests show that the full Newton minimization method achieves a high convergence rate and a reasonably accurate reconstruction of the model parameters. Taking advantage of a direct solver based on LU decomposition, the full Newton minimization method can also be implemented in a matrix-free manner. Tests with this algorithm were conducted with the BP/EAGE velocity model and highlight its high performance capabilities. To our knowledge, this is the first implementation of FWI with simultaneous sources using the full Newton method on a large scale model.
3D Inversion of DC/IP Data Using Adaptive OcTree Meshes
Haber, E. (University of British Columbia) | Oldenburg, D. (University of British Columbia) | Shekhtman, R. (University of British Columbia) | Granek, J. (Computational Geosciences Incorporated) | Marchant, D. (Computational Geosciences Incorporated) | Holtham, E. (Computational Geosciences Incorporated)
SUMMARY Data acquired from a direct current (DC) and induced polarization (IP) survey can be used to recover the conductivity and chargeability structures of the subsurface of the earth. In order to maximize the value of such a survey, the data should be inverted in 3D. As surveys get larger and targets get more complex, the discretization applied in regular rectilinear meshes can become cumbersome, resulting in prohibitively large numbers of cells. This problem is exacerbated in the presence of severe topography, or in cases of irregular survey geometry. Applying an adaptive OcTree mesh structure, it is possible to obtain fine resolution cells in regions of high variability without adding unnecessarily small cells where they are not required. This results in a vastly decreased number of cells, without penalizing the potential for high resolution recovered models. We develop the DC/IP inverse algorithm on the OcTree mesh, and apply it to a field example from South America.
- South America (0.25)
- North America (0.15)
Summary The time-lapse OBN (Ocean Bottom Node) seismic data acquired by Shell in 2007 and 2010 in the Mars field, is a true wide azimuth marine dataset that provides a good opportunity to study the azimuth and offset dependence of time-lapse depth-shifts and amplitudes. The goals of such analysis are to gain a better understanding of the response of seismic waves to geomechanical changes happening in the field and to improve our ability to discriminate between pressure and saturation signals using the time-lapse AVO response. In this paper we compare the conventional time-lapse AVO analysis with an alternative method based on the combined use of time-lapse amplitudes and time-lapse depth-shifts to separate pressure and fluid-related changes on the Mars time-lapse OBN data.
- North America > United States > Gulf of Mexico > Central GOM > East Gulf Coast Tertiary Basin > Mississippi Canyon > Block 851 > Mars Field (0.99)
- North America > United States > Gulf of Mexico > Central GOM > East Gulf Coast Tertiary Basin > Mississippi Canyon > Block 850 > Mars Field (0.99)
- North America > United States > Gulf of Mexico > Central GOM > East Gulf Coast Tertiary Basin > Mississippi Canyon > Block 808 > Mars Field (0.99)
- (5 more...)
Summary A multicomponent (4C) towed-streamer that measures not only scalar pressure wavefields, but also the three components of P-wave particle motion, was proposed by Robertsson et al. (2008). These new measurements may be used to perform joint interpolation and deghosting of the pressure wavefield, and open the possibility to reconstruct the seismic wavefield at any point between streamers using both pressure and the crossline component of the pressure gradient. To be able to put these proposed applications in practice, good fidelity of the 4C measurement is needed. We study the effects of the streamer mechanical design on the signal fidelity, review the sensor requirements, and present data examples from a 4C multicomponent streamer, demonstrating the possibility to record high signal fidelity in such a 4C multicomponent towed streamer.
Contributions of the Horizontal and Vertical Components of Particle Velocity in 3D Pressure Wavefield Reconstruction on Dense Receiver Grids Using Generalized Matching Pursuit
Vassallo, Massimiliano (WesternGeco) | Eggenberger, Kurt (WesternGeco) | van Manen, Dirk-Jan (WesternGeco) | Özdemir, Kemal (WesternGeco) | Robertsson, Johan (ETH Zürich) | Özbek, Ali (Schlumberger)
Summary Multichannel 3D reconstruction and deghosting techniques based on multicomponent streamer measurements of the pressure wavefield and its associated gradients were recently introduced in literature. In particular, the Generalized Matching Pursuit (GMP) technique was applied to multicomponent 3D synthetic data bringing significant improvements to address the aliasing arising from sparse crossline sampling. In this abstract, we present an example of real data acquired by an experimental 3D towed multicomponent cable array and show the performance of GMP applied to the multicomponent measurements. The real data examples illustrate that GMP reconstructs and deghosts the pressure wavefield onto a 2D receiver grid uniformly sampled at 6.25 m in both, the inline and the crossline directions, starting from a very limited number of crossline samples at realistic spacings (i.e., 75 m). We analyze the contribution of each component to the overall crossline reconstruction. We show that the crossline component of particle velocity is the key enabler for GMP to produce a very effective and robust reconstruction of the three-dimensional wavefield back-scattered by the subsurface for each recorded seismic shot.
- North America > United States (0.16)
- Europe (0.16)