... the directional information of the ray. So the stereotomography has a angles are no longer used as the model parameters, and the natural advantage in inverting anisotropic parameters. Stereotomography is a data domain oriented kinematic model space is correspondingly reduced to three, the number tomography method...

... compliance with model space used for training and do not enable the network the requirement of physics for the wave propagation. The to evolve and adapt toward different data distributions domains of physics driven (Phy) optimization, based on data previously unseen by the network. misfit functionals...

...Interrogating the model space of airborne electromagnetic inversion to answer a hydrogeologic question Seogi Kang1*, Rosemary Knight1, and Todd J. Greene2 1Stanford University, CA, U.S.A., 2California State University, Chico, CA, U.S.A. Summary method, and quantified associated uncertainties...

... is described as follow: Introduction 1) Define the prior model space, and sample In geophysics, assessing the uncertainty on a given model by models in using field data is a core issue. It is classically performed via Gaussian noise propagation in the model (e.g., Kemna et al., 2) Generate the model responses...

... reduction artificial intelligence dimensionality salt structure stochastic fwi model space deep autoencoder stochastic full waveform inversion neural network global optimization inversion waveform inversion velocity model machine learning long-wavelength velocity model initial model optimal...

... with the physics of the forward problem to estimate the aposteriori probability density function in the model space. The probability distribution derived with this approach can be used to quantify the uncertainty in the estimation of vertical resistivity profile. We apply our inversion scheme on three synthetic...

... wavepath b-spline projection upstream oil & gas model space albertin full waveform inversion inversion artificial intelligence elastic full waveform inversion equation phase rotation objective function annual international meeting Clement , F. , G. Chavent , and S...

... simplification hessian artifact mapping migration inversion annual meeting reservoir characterization seg international exposition linearized inversion modeling operator lsmig upstream oil & gas kirchhoff migration model space noise operator Claerbout , J. F. , 1971 , Toward...

... stereotomography algorithm derivative matrix model space iteration structure tensor reflector position fréchet derivative matrix annual meeting information geological horizon tomography seg seg international exposition geological framework regularization technique upstream oil...

... the challenges of the proposed method. Presentation Date: Monday, September 25, 2017 Start Time: 4:20 PM Location: 340A Presentation Type: ORAL application velocity model caumon artificial intelligence guillaume caumon data space model space space misfit seismic data macro-layered velocity...

... space. workflow noise inversion result application full waveform inversion seismic reflection data virieux gas sand geophysics inversion upstream oil & gas information reservoir characterization gradient brossier laplacian model space fwi gradient full-waveform inversion...

... Summary In this letter, a new inversion method called successive differential evolution (DE-S) algorithm was proposed to estimate Q factor and velocity from the zero-offset VSP record. The method is suitable for the high-dimension nonseparable model space. We developed the bottom-up...

... this general methodology, we apply an EM inversion to a suite of models in one, two, and three dimensions, and suggest the importance of choos- ing an appropriate model space based on the goal of the EM inversion. INTRODUCTION Electromagnetic (EM) fields and fluxes can be used to excite the earth...

... Summary In this paper, we combine a fast wave equation solver using boundary integral methods with a global optimization method, namely Particle Swarm Optimization (PSO), to estimate an initial velocity model. Unlike finite difference methods that discretize the model space into pixels...

... Summary Monte Carlo optimizations are based on iteratively computing residuals through forward modeling. Forward modeling thousands of times can be computationally expensive, particularly for large model spaces. We present a highly efficient 2D finite element gravity model that will be part...

... in accuracy and computational efficiency can be obtained in wavefield separation. Examples from the synthetic data and real borehole seismic data (both VSP data and cross-well seismic data) verify this method is feasible and valid. seg houston 2013 synthetic data model space linear radon transform...

... the Hessian is due to the fact that it is too large to be represented as a matrix, and applying it to a function in model space is in itself a very computationally intensive operation. To leading order, the wave-equation Hes- sian is the composition F F , where F is the migration op- erator, and F...

... strategy realization correspond equation multimodal problem stochastic algorithm model space artificial intelligence algorithm reflectivity term travel time cmae upstream oil & gas different stochastic algorithm main menu stochastic algorithm base trace inversion Solving...

... resistivities acquired with induction or laterolog instruments. Two field cases of application verify the accuracy and stability of the new Bayesian inversion method. inverse problem resistivity log model space borehole resistivity measurement bayesian inversion method resistivity measurement...

...) operators, the former mapping from model space to data space, and the latter mapping from data space to model space. SPDR2 uses least-squares migration to find an optimal model which, in turn, is mapped to data space using de-migration, providing a reconstructed shot gather. We apply SPDR2 to real data from...