Disclosed are systems and methods for numerically simulating seismic-wave propagation in tilted transversely isotropic (TTI) media, using an adaptive Lebedev staggered grid. In various embodiments, the adaptive grid includes multiple horizontal zones having different associated grid spacings, which may be determined based on a vertical wave-velocity model. The numerical simulation may involve iteratively solving a set of finite-difference equations including finite-difference coefficients that vary spatially depending on the grid spacing. Additional embodiments and features are described.

Methods and systems for separating seismic data acquired using a plurality of substantially simultaneously fired sources are described. The sources use sweep sequences having low cross correlation levels to generate seismic waves, and their source signatures are determined. Using the source signatures, the wave fields associated with each of the sources are extracted from the seismic data by, for example, performing a time domain deconvolution.

Methods for separating the unknown contributions of two or more sources from a commonly acquired set of wavefield signals based on varying parameters at the firing time, location and/or depth of the individual sources in a lateral 2D plane.

Disclosed herein is a method of stripping a strong reflection layer based on deep learning. The method establishes a direct mapping relationship between a strong reflection signal and seismic data of a target work area through a nonlinear mapping function of the deep neural network, and strips a strong reflection layer after the strong layer is accurately predicted. A mapping relationship between the seismic data containing the strong reflection layer and an event of the strong reflection layer is directedly found through training parameters. In addition, this method does not require an empirical parameter adjustment, and only needs to prepare a training sample that meets the actual conditions of the target work area according to the described rules.

A method of seismic modeling using an elastic model, the elastic model including a grid having a grid spacing sized such that, when synthetic seismic data is generated using the elastic model, synthetic shear wave data exhibits numerical dispersion, the method including: generating generated synthetic seismic data using the elastic model, wherein the generated synthetic seismic data includes synthetic compression wave data and synthetic shear wave data, and modifying the generated synthetic seismic data to produce modified synthetic seismic data by attenuating at least some of the synthetic shear wave data in order to attenuate at least some of the numerically dispersive data.

A method can include selecting a type of geophysical data; selecting a type of algorithm; generating synthetic geophysical data based at least in part on the algorithm; training a deep learning framework based at least in part on the synthetic geophysical data to generate a trained deep learning framework; receiving acquired geophysical data for a geologic environment; implementing the trained deep learning framework to generate interpretation results for the acquired geophysical data; and outputting the interpretation results.

A method of compressing data from seismic waves using Gabor frames utilizes a plurality of geophones positioned within a region of interest. Each of the plurality of geophones is communicably coupled with at least one remote server. Thus, a plurality of reflected-seismic signals received through the plurality of geophones can be transmitted to the at least one remote server for analyzing and calculations. The plurality of reflected-seismic signals is converted into a set of Gabor frames, wherein the Gabor frames is generated via a plurality of prolate spheroidal wave functions (PSWF). A Gabor frame-generating calculation module utilizes the plurality of PSWF to generate the set of Gabor frames. A dual frame for each of the set of Gabor frames is derived and used for quantization purposes. Preferably, a tree structured vector quantization process is followed.

Disclosed herein is a method of stripping a strong reflection layer based on deep learning. The method establishes a direct mapping relationship between a strong reflection signal and seismic data of a target work area through a nonlinear mapping function of the deep neural network, and strips a strong reflection layer after the strong layer is accurately predicted. A mapping relationship between the seismic data containing the strong reflection layer and an event of the strong reflection layer is directedly found through training parameters. In addition, this method does not require an empirical parameter adjustment, and only needs to prepare a training sample that meets the actual conditions of the target work area according to the described rules.

A method can include receiving microseismic data of microseismic events as acquired by sensors during hydraulic fracturing of a geologic region; jointly calibrating sensor orientation of the sensors and a velocity model of the geologic region via an objective function and the microseismic data; and, based at least in part on the jointly calibrating, determining one or more locations of the one or more microseismic events.

A method of compressing data from seismic waves using Gabor frames utilizes a plurality of geophones positioned within a region of interest. Each of the plurality of geophones is communicably coupled with at least one remote server. Thus, a plurality of reflected-seismic signals received through the plurality of geophones can be transmitted to the at least one remote server for analyzing and calculations. The plurality of reflected-seismic signals is converted into a set of Gabor frames, wherein the Gabor frames is generated via a plurality of prolate spheroidal wave functions (PSWF). A Gabor frame-generating calculation module utilizes the plurality of PSWF to generate the set of Gabor frames. A dual frame for each of the set of Gabor frames is derived and used for quantization purposes. Preferably, a tree structured vector quantization process is followed.