... measurements and potential overfitting in complex geological settings. Our methodology involves preprocessing UHRS and CPT data, interpreting and integrating geological unit (GU) information, and applying repeated k-fold cross-validation to evaluate model performance. The field data results demonstrate...

... repercussions without promoting the energy transition. While more and more States will seek to use environmental issues to justify certain questionable decisions, it is important that all citizens have access to a fair vision of the situation, and that they are well informed of the real impacts as well...

... The precise lithologic reservoirs spatial characterization isstill a challenging task because of strong heterogeneity, unclear sand connectivity and unclear pinchout. However, in terms of identifying the effectiveness of lithologic trap, reservoir prediction based on amplitude information have...

... preconditioning scheme to provide a better sampled initial model for DL interpolation. The DL model works iteratively, passing the data between the unsupervised learning architecture for feature extraction and the iterative framework for reconstruction. An attention network highlights the important information...

.... In this study, we depart from conventional stacking approaches. Instead, we train symmetric autoencoders (SymAE) to learn the capability of disentangling subsurface information, which is coherent across multiple noise gathers, from the remaining source effects. We focus on a controlled experiment aimed...

... Estimating the seismic impedance from the band-limited seismic data has proven to be a difficult task, especially due to the absence of low frequencies. This situation can be mitigated using the information obtained from well logs which can be used as a constraint. This prior information can...

... in Geothermal Energy (FORGE) project in Utah, direct location of microseismic events stimulated from fluid injection can offer vital insights about fracture propagation and reservoir connectivity for more informed production decisions. Surface and borehole seismic arrays are usually deployed for monitoring...

... Conventional marine seismic surveys lack near offset information, resulting in lower-quality imaging in the shallow subsurface. We present a deep learning workflow to reconstruct near offsets, which utilizes field data collected with source-over-cable acquisition geometry as training data. First...

... Wide-azimuth seismic is the basis of pre-stack seismic anisotropy prediction. The conventional pre-stack depth migration method is based on the ground observation system. The output gather is offset gather, which lacks valuable azimuth information. The OVT domain processing method can output...

.... Although the spectrum energy of SSMP can be gathered to the center time and frequency of the selected matching wavelet. However, there will be a phenomenon of over-compression. A lot of effective formation information will be eliminated, resulting in unreal results and false energy with high resolution...

... geologist geological subdiscipline flexure exploration geophysicist reservoir characterization potential field applied geoscience & energy 10 ambiguity altitude inversion gravity data deconvolution interpretation geophysical prospecting information geophysics application Assessing...

... geoscience & energy 10 quantitative requirement geologist artificial intelligence human-in-the-loop ai llm objective american association energy industry annotation data annotation arxiv university deep learning machine learning exploration geophysicist information threshold neural...

... on reservoir performance, and the challenges in predicting reservoir behavior due to internal heterogeneities underscore the need for detailed 3D facies distribution models. The integration of geological concepts, analog information, and process-based models offers a sophisticated approach to enhancing...

... information of the dictionary and representation coefficients, enhancing the flexibility of the convolutional dictionary and coefficients. The advantages of this approach include: leveraging convolutional sparse coding to decompose images fully, considering the local correlation in seismic data, which...

... (GOM) shows that the Bayesian method provides not only reasonable estimates of pressure and fluid saturation changes but also comprehensive uncertainty information for decision making and risk assessment. It is worth noting that the quality of machine learning estimates is important and should...

... characterization vpvs ratio geologist geomechanics inversion artificial intelligence economic geology vp model information density inversion peng shen depth profile applied geoscience & energy 10 hydrophone data high quality sand exploration geophysicist geo-statistical inversion petroleum...

... geological subdiscipline structural geology strike-slip fault machine learning information deep learning cascade ant american association strike-slip fault skeletonization ant tracking fracture characterization rock type reservoir characterization ant fault identification accuracy applied...

... transformation technique for the high-angle well situation. The former technique obtains the configuration and orientation of the geological structure around the borehole; the latter technique, using the borehole trajectory information, allows us to view the structure image in the earth coordinates system...

... an accurate velocity model is a key step that directly impacts the subsequent interpretation. Wave-equation reflection traveltime inversion (WRTI) is an effective method that can adapt to complex structures and utilize seismic reflection traveltime information to establish deep subsurface velocity models...

... The construction of the prior model is an important part of the stochastic inversion method. In order to obtain reliable and reasonable prior information about the inversion parameters, we proposed a method to construct the Von Karman random medium prior model. Compared with the traditional...