Abstract: A method for reconstructing at least one trace in a seismic image of a common receiver and time domain includes a convolutional neural network trained under an unsupervised learning approach with a modified receptive field.

Abstract: Systems, computer-readable media, and methods are provided. Blended baseline data is generated by numerically blending unblended baseline data according to a simultaneous shooting schedule scheme. Pseudo-deblended baseline seismic data is generated by applying a pseudo-deblending procedure to the blended baseline data. Machine learning labels are generated from common gathers of the pseudo-deblended baseline data and the unblended baseline data. A neural network is trained using the labels, the common gathers of the pseudo-deblended baseline data, and the unblended baseline data to produce common gathers of deblended baseline seismic data from the common gathers of the pseudo-deblended baseline seismic data.

Abstract: A method includes receiving input including baseline data representing a subsurface volume prior to an injection operation, injection data representing an injection operation during a first timestep, and an initial pressure and saturation data at the beginning of the first timestep, training a machine learning model to predict a first pressure and saturation map at an end of the first timestep based on baseline data, the injection data, and the initial pressure and saturation data, and training the machine learning model to predict a second pressure and saturation model at an end of a second timestep based on the baseline data, injection data representing the injection operation during the second timestep, and the first pressure and saturation map at the end of the first timestep. The trained machine learning model is configured to predict an implementation pressure and saturation map at a plurality of times during an injection operation.

Abstract: A method for modeling a subsurface volume using time-lapse data includes receiving a baseline seismic dataset, a baseline property model, a monitoring seismic dataset. and a monitoring property model. sorting the baseline seismic dataset and the monitoring seismic dataset into respective common gathers, representing offset, time, and depth point, extracting signal data for a range of depth points for the baseline dataset and a signal data for a corresponding range of depth points for the monitoring seismic dataset, predicting a property model change based at least in part on the signal data for the range of depth points of the baseline seismic dataset and the monitoring seismic dataset, using a machine learning model, and generating a property model representing a subsurface volume based at least in part on the property model change predicted using the machine learning model.

Abstract: A method for seismic surveying includes receiving a baseline dataset and a plurality of sparse monitoring datasets, generating a decimated baseline dataset by removing one or more sources, receivers, or both from the baseline dataset, generating a reconstructed baseline dataset by inputting the decimated baseline dataset into a machine learning model, generating reconstructed monitoring datasets by inputting the plurality of sparse monitoring datasets to the machine learning model, the machine learning model having been trained based on a comparison of the reconstructed baseline dataset to the baseline seismic dataset, determining accuracies for the plurality of sparse monitoring datasets by comparing the reconstructed monitoring datasets to the baseline dataset, and selecting one or more survey geometries for arranging physical sources and physical receivers in a seismic survey based at least in part on the accuracies of the plurality of sparse monitoring datasets.

Abstract: Systems, computer-readable media, and methods are provided. Blended baseline data is generated by numerically blending unblended baseline data according to a simultaneous shooting schedule scheme. Pseudo-deblended baseline seismic data is generated by applying a pseudo-deblending procedure to the blended baseline data. Machine learning labels are generated from common gathers of the pseudo-deblended baseline data and the unblended baseline data. A neural network is trained using the labels, the common gathers of the pseudo-deblended baseline data, and the unblended baseline data to produce common gathers of deblended baseline seismic data from the common gathers of the pseudo-deblended baseline seismic data.

Abstract: An integrated workflow is presented including a suite of data-driven technologies that aims to substantially reduce the cost of monitoring data acquisition, improve the robustness and efficiency of time-lapse data processing procedures to shorten the turnaround time of projects utilizing seismic data for monitoring sub-surface fluid reservoirs. In particular, plumes of subsurface CO2 may be monitored, including CO2 deliberately injected into the sub-surface as a sequestration technique. The workflow may include two parts: (1) cost-effective data acquisition schemes and (2) efficient data processing algorithms. The technology components in the workflow may include deep learning sparse monitoring data reconstruction and optimal acquisition survey design, deep learning deblending of simultaneous source monitoring data, time-lapse data repeatability enforcement through deep learning, and rapid CO2 plume body and property estimation directly from pre-migration monitoring data.

Abstract: The present invention is related to a method for reconstructing at least one trace in a seismic image of a common receiver and time domain, the image comprising traces in time domain with seismic data and one or more traces to be reconstructed. A first aspect of the invention is a method that is characterized by a specific use of a convolutional neural network trained under an unsupervised learning approach with a modified receptive field. A second aspect of the invention is a deblending method based on the use of a reconstructing method according to the first aspect of the invention applied to a denoising step of a deblending process allowing a very effective data acquisition while keeping a high quality output data sets after being processed according to the first and/or second aspects of the invention.

Abstract: A computer-implemented method for obtaining reconstructed seismic data for determining a subsurface feature, includes: determining an initial training velocity model, training a machine learning model based on first training seismic data and second training seismic data generated from the training velocity model, the first training seismic data corresponding to one or more first frequencies, the second training seismic data corresponding to one or more second frequencies lower than the one or more first frequencies, obtaining, based on measured seismic data and the machine learning model, reconstructed seismic data corresponding to the one or more second frequencies, generating a velocity model based on the measured seismic data, the reconstructed seismic data, and a full waveform inversion (FWI), and when the generated velocity model does not satisfy a preset condition, updating the training velocity model based on the generated velocity model, to obtain updated reconstructed seismic data for determining a s

Abstract: A computer-implemented method for determining a subsurface feature, includes: determining a first velocity model based on an initial velocity model; generating a second velocity model based on measured seismic data at one or more first frequencies, the first velocity model, and a full waveform inversion (FWI); and in response to the second velocity model not satisfying a preset condition, performing a seismic forward simulation on the second velocity model to generate simulated seismic data at one or more second frequencies lower than the one or more first frequencies; updating the first velocity model based on the simulated seismic data at the one or more second frequencies; and updating the second velocity model based on the measured seismic data at the one or more first frequencies, the updated first velocity model, and the FWI, to determine the subsurface feature.

Abstract: A computer-implemented method for determining a subsurface feature, includes: determining a first velocity model based on an initial velocity model; generating a second velocity model based on measured seismic data at one or more first frequencies, the first velocity model, and a full waveform inversion (FWI); and in response to the second velocity model not satisfying a preset condition, performing a seismic forward simulation on the second velocity model to generate simulated seismic data at one or more second frequencies lower than the one or more first frequencies; updating the first velocity model based on the simulated seismic data at the one or more second frequencies; and updating the second velocity model based on the measured seismic data at the one or more first frequencies, the updated first velocity model, and the FWI, to determine the subsurface feature.

Abstract: A computer-implemented method for obtaining reconstructed seismic data for determining a subsurface feature, includes: determining an initial training velocity model, training a machine learning model based on first training seismic data and second training seismic data generated from the training velocity model, the first training seismic data corresponding to one or more first frequencies, the second training seismic data corresponding to one or more second frequencies lower than the one or more first frequencies, obtaining, based on measured seismic data and the machine learning model, reconstructed seismic data corresponding to the one or more second frequencies, generating a velocity model based on the measured seismic data, the reconstructed seismic data, and a full waveform inversion (FWI), and when the generated velocity model does not satisfy a preset condition, updating the training velocity model based on the generated velocity model, to obtain updated reconstructed seismic data for determining a s

Abstract: A full wave inversion (FWI) may utilize an Amplitude-Frequency-Differentiation (AFD) or a Phase-Frequency-Differentiation (PFD) operation to form a velocity model of a subterranean formation utilizing recovered low wavenumber data. Received seismic data is processes to isolate two data signals at different frequencies. In an AFD operation, the two data signals are summed and the data of the envelope of the summed function is used for the FWI. In a PFD operation, the phase data of the quotient of the two data signals is used for the FWI. The FWI proceeds iteratively utilizing either the AFD or PFD data or with single frequency data until the cost function of the AFD or PFD is satisfied.

Abstract: A memory-efficient Q-RTM computer method and apparatus for imaging seismic data is described. A seismic image may be formed from a memory-efficient Q-RTM module utilizing received attenuated seismic data. Seismic data is processed by the memory-efficient Q-RTM module to compensate for amplitude attenuation and phase velocity dispersion simultaneously during back-propagation in RTM. A negative quality factor, Q, is obtained by modifying the wave equation to compensate for amplitude attenuation. One or more dispersion optimization terms introduced to a wave equation for compensation of Q effects on the phase, solved by a finite difference algorithm, compensate for phase velocity change and further adjust amplitude attenuation compensation.

Abstract: Method for reconstructing subsurface Q depth profiles from common offset gathers (92) of reflection seismic data by performing migration (40), ray tracing (100), CDP-to-surface takeoff angle finding (96, 98), kernel matrix construction (110), depth-to-time conversion and wavelet stretching correction (80), source amplitude spectrum fitting, centroid frequency shift calculation (90), and box-constrained optimization (120).

Abstract: A memory-efficient Q-RTM computer method and apparatus for imaging seismic data is described. A seismic image may be formed from a memory-efficient Q-RTM module utilizing received attenuated seismic data. Seismic data is processed by the memory-efficient Q-RTM module to compensate for amplitude attenuation and phase velocity dispersion simultaneously during back-propagation in RTM. A negative quality factor, Q, is obtained by modifying the wave equation to compensate for amplitude attenuation. One or more dispersion optimization terms introduced to a wave equation for compensation of Q effects on the phase, solved by a finite difference algorithm, compensate for phase velocity change and further adjust amplitude attenuation compensation.

Abstract: A full wave inversion (FWI) may utilize an Amplitude-Frequency-Differentiation (AFD) or a Phase-Frequency-Differentiation (PFD) operation to form a velocity model of a subterranean formation utilizing recovered low wavenumber data. Received seismic data is processes to isolate two data signals at different frequencies. In an AFD operation, the two data signals are summed and the data of the envelope of the summed function is used for the FWI. In a PFD operation, the phase data of the quotient of the two data signals is used for the FWI. The FWI proceeds iteratively utilizing either the AFD or PFD data or with single frequency data until the cost function of the AFD or PFD is satisfied.

Abstract: A full wave inversion (FWI) may utilize an Amplitude-Frequency-Differentiation (AFD) or a Phase-Frequency-Differentiation (PFD) operation to form a velocity model of a subterranean formation utilizing recovered low wavenumber data. Received seismic data is processes to isolate two data signals at different frequencies. In an AFD operation, the two data signals are summed and the data of the envelope of the summed function is used for the FWI. In a PFD operation, the phase data of the quotient of the two data signals is used for the FWI. The FWI proceeds iteratively utilizing either the AFD or PFD data or with single frequency data until the cost function of the AFD or PFD is satisfied.

Abstract: Method for reconstructing subsurface Q depth profiles from common offset gathers (92) of reflection seismic data by performing migration (40), ray tracing (100), CDP-to-Data Domain surface takeoff angle finding (96, 98), kernel matrix construction (110), depth-to-time conversion and wavelet stretching correction (80), source amplitude spectrum fitting, centroid frequency shift calculation (90), and box-constrained optimization (120).

Abstract: Method for reconstructing subsurface profiles for seismic velocity or other geophysical properties from recorded seismic data. In one embodiment, a starting model of seismic velocity is assumed (10). The computational domain is divided into two (or more) subdomains by horizontal planes based on an analysis of velocity model (30), and the allowed maximum grid size for each subdomain is determined (50). Auxiliary perfectly matched layers (PML's) are attached to each planar interface between subdomains (80), e.g. two PML's on each side of the interface between the coarse and fine subdomains. Simulated seismic data are computed using the SG-DO technique (100-230). The simulated seismic data are compared to the recorded seismic data, then the residual is calculated (240) and used to update the model (320). The method may be iterated until the model is suitably converged (260).