A distributed acoustic sensor is positioned within a wellbore of a geologic formation. Seismic waves are detected using the distributed acoustic sensor. A raw seismic profile is generated based on the detected seismic waves. Resonant noise is identified and reduced in seismic data associated with the raw seismic profile.

A method for estimating a time variant signal representing a seismic source obtains seismic data recorded by at least one receiver and generated by the seismic source, the recorded seismic data comprising direct arrivals and derives the time variant signal using an operator that relates the time variant signal to the acquired seismic data, the operator constrained such that the time variant signal is sparse in time.

A method for correction of a sea surface state can include receiving geophysical data from a seismic survey, wherein the seismic survey utilizes a plurality of receivers disposed in a body of water and at least one source in the body of water, actuated at a plurality of shot points. The method can include identifying, in the geophysical data, a wavefield based on the actuation of the at least one source, and determining, based on the identified wavefield, a sea surface state at the at least one source at one of the plurality of shot points.

A method for use in seismic exploration comprises: accessing a set of seismic data representative of a subterranean geological formation and a subsurface attribute model of the subterranean geological formation; performing a wavefield extrapolation on the seismic data in the subsurface attribute model; applying the time-shift extended imaging condition to the extrapolated wavefields; forming shot-indexed, time shift gathers for each image pixel of the subsurface attribute model from the conditioned extrapolated wavefields; adaptively focusing the gathers; and stacking the adaptively focused gathers; and imaging the subterranean geological formation from the stacked, adaptively focused gathers. The method may, in some aspects, be realized by a computing apparatus programmed to perform the method or as a set of instructions encoded on a non-transitory program storage medium that, when executed by a computing apparatus, perform the method.

Techniques are disclosed relating to machine learning in the context of noise filters for sensor data, e.g., as produced by geophysical surveys. In some embodiments, one or more filters are applied to sensor data, such a harsh filter determined to cause a threshold level of distortion in measured reflections, a mild filter determined to leave a threshold level of remaining noise signals, or an acceptable filter. In some embodiments, the system trains a machine learning classifier based on outputs of the filtering procedures and uses the classifier to determine whether other filtered sensor data from the same survey exhibits acceptable filtering. This may improve accuracy or performance in detecting unacceptable filtering, in some embodiments.

A method includes acquiring blended seismic data representing a subsurface volume of interest from a plurality of seismic sources, estimating a signal mode using one or more first priors by applying sparse inversion to the blended seismic data, predicting multi-source interference in the blended seismic data based at least in part on the estimated signal mode, removing the estimated signal mode and the predicted multi-source interference from the blended seismic data, such that a residual signal is left, and estimating a coherent signal from the residual signal by solving a sparse inversion.

Systems and methods for performing focused blind deconvolution of signals received by a plurality of sensors are disclosed. In some embodiments, this may include determining a cross-correlation of first and second signals, obtaining a cross-correlation of a first response function and a second response function based on the cross-correlation of the first and second signals and subject to a first constraint that the first and second response functions are maximally white, and obtaining the first and second response functions based on the cross-correlation of the first and second response functions and subject to a second constraint that the first and second response functions are maximally front-loaded.

A method of enhancing seismic images includes receiving a seismic gather. The seismic gather includes a plurality of seismic traces. A feature trace is generated based on the plurality of seismic traces in the seismic gather. For each of the plurality of seismic traces in the seismic gather, a correlation trace is generated based on that seismic trace and the feature trace, the correlation trace is modified using an activation function, and an enhanced trace is generated by multiplying that seismic trace with the modified correlation trace.

Disclosed is a method of, and computer program and apparatus for, identifying reflected signals, subsequent to their reflection within a medium. The method comprises obtaining return signals (100), resulting from measurements being performed over a measurement period. The measurement period comprises sub-periods, the return signals comprising reflected signals and noise. The plurality of return signals are partitioned into plural sets (220) of equal cardinality or as equal as possible such that their cardinality differs by no more than one. A stacked correlation value is determined (130) for the return signals by determining the mean of the return signals across the plural sets (230) and determining a correlation value of the plural sets over each of the time sub-periods (240). Peaks in the variation of the stacked correlation value over time can then be identified and each of the peaks in the variation of the stacked correlation value over time can be attributed to a reflected signal.

A method for estimating a time variant signal representing a seismic source obtains seismic data recorded by at least one receiver and generated by the seismic source, the recorded seismic data comprising direct arrivals and derives the time variant signal using an operator that relates the time variant signal to the acquired seismic data, the operator constrained such that the time variant signal is sparse in time.