A method can include receiving seismic survey data of a subsurface environment from a seismic survey that includes a source arrangement of sources that is spatially denser than a receiver arrangement of receivers; processing the seismic survey data using the principle of reciprocity for performing interpolation across the receivers to generate processed seismic survey data; and generating an image of at least a portion of the subsurface environment using the processed seismic survey data.

Systems and methods for filtering interface waves from single component seismic data are disclosed. In one embodiment, a method of filtering seismic data includes comparing amplitude coefficients of a matrix storing the seismic data in a time-frequency domain against an amplitude threshold, and comparing frequencies of the matrix against a maximum expected frequency of noise. The method further includes, for each amplitude coefficient having less than the amplitude threshold and an associated frequency less than the maximum expected frequency of noise, scaling the amplitude coefficient to reduce its value. The method also includes performing an inverse time-frequency transformation on the matrix to generate a noise model in a time domain, and subtracting the noise model from the seismic data in the time domain to generate filtered seismic data.

A method includes receiving well log data comprising a plurality of well logs, identifying one or more sections of one or more well logs of the plurality of well logs that have substantially complete data, training a reconstruction neural network to reconstruct incomplete well logs based on the one or more sections of the one or more well logs that have substantially complete data, and reconstructing one or more incomplete well logs of the plurality of well logs using the reconstruction neural network.

Seismic image processing including filtering a three-dimensional (3D) seismic image for random noise attenuation via multiple processors. The filtering includes receiving a 3D image cube of seismic image data, decomposing the 3D image cube into 3D sub-cubes for parallel computation on the multiple processors, designing and applying a two-dimensional (2D) adaptive filter for image points on 2D image slices of the 3D sub-cubes via the multiple processors to give filtered 3D sub-cubes, and summing the filtered 3D sub-cubes to give a filtered 3D image cube.

Waves from cement bond logging with a sonic logging-while-drilling tool (LWD-CBL) are often contaminated with tool waves and may yield biased CBL amplitudes. The disclosed LWD-CBL wave processing corrects the first echo amplitudes of LWD-CBL before calculating the BI. The LWD-CBL wave processing calculates a tool wave amplitude and a phase angle difference as the difference of the phases between the tool waves and casing waves. The tool waves are then used to correct the LWD-CBL casing wave amplitude and remove errors introduced from tool waves. In conjunction with the sets of operations described, the LWD-CBL wave processing also include array preprocessing operations. Array preprocessing may employ variation of bandpass filtering and frequency-wavenumber (F-K) filtering operations to suppress tool wave.

A method for enhancing properties of geophysical data with deep learning networks. Geophysical data may be acquired by positioning a source of sound waves at a chosen shot location, and measuring back-scattered energy generated by the source using receivers placed at selected locations. For example, seismic data may be collected using towed streamer acquisition in order to derive subsurface properties or to form images of the subsurface. However, towed streamer data may be deficient in one or more properties (e.g., at low frequencies). To compensate for the deficiencies, another survey (such as an Ocean Bottom Nodes (OBN) survey) may be sparsely acquired in order to train a neural network. The trained neural network may then be used to compensate for the towed streamer deficient properties, such as by using the trained neural network to extend the towed streamer data to the low frequencies.

A method for enhancing properties of geophysical data with deep learning networks. Geophysical data may be acquired by positioning a source of sound waves at a chosen shot location, and measuring back-scattered energy generated by the source using receivers placed at selected locations. For example, seismic data may be collected using towed streamer acquisition in order to derive subsurface properties or to form images of the subsurface. However, towed streamer data may be deficient in one or more properties (e.g., at low frequencies). To compensate for the deficiencies, another survey (such as an Ocean Bottom Nodes (OBN) survey) may be sparsely acquired in order to train a neural network. The trained neural network may then be used to compensate for the towed streamer deficient properties, such as by using the trained neural network to extend the towed streamer data to the low frequencies.

Methods and systems for removing tool mode waveforms. The method may include disposing a bottom hole assembly (BHA) into a wellbore. The BHA may comprise at least one transmitter configured to transmit a pressure pulse and at least one receiver configured to record one or more waveforms. The method may further comprise performing a logging-while-drilling (LWD) operation in which the one or more waveforms are recorded with the at least one receiver, transmitting the one or more waveforms to an information handling system, removing one or more tool mode waveforms from the one or more waveforms to form an updated set of waveforms, and forming differential phase time semblance map based at least in part on the updated set of waveforms. The system may comprise the BHA and information handling system configured to remove one or more tool mode waveforms.

Seismic data is deblended by performing, for each receiver, a first inversion and a second inversion in a transform domain. The first inversion is formulated to minimize a number of non-zero coefficients of the first inversion result. A sub-domain of the transform domain is defined by vectors of a transform domain basis for which the first inversion has yielded the non-zero coefficients. The second inversion is performed in this sub-domain. The solution of the second inversion is used to extract deblended seismic datasets corresponding to each of the distinct signals, from the seismic data.

A machine-implemented method, at least one non-transitory computer-readable medium storing instructions, and a computing system are provided for attenuating noise. A computing system receives a seismic image and generates a first image using a first neural network configured to identify low-frequency ground roll in a seismic image, and a second image using a second neural network configured to identify reflections in the seismic image. A combined image is generated by combining the first image and the second image. The first neural network and the second neural network are adjusted to reduce a difference between the combined image and the seismic image using frequency constraint to guide separation of the seismic image into the first image and the second image.