This disclosure describes a system and method for generating a subsurface model representing lithological characteristics and attributes of the subsurface of a celestial body or planet. By automatically ingesting data from many sources, a machine learning system can infer information about the characteristics of regions of the subsurface and build a model representing the subsurface rock properties. In some cases, this can provide information about a region using inferred data, where no direct measurements have been taken. Remote sensing data, such as aerial or satellite imagery, gravimetric data, magnetic field data, electromagnetic data, and other information can be readily collected or is already available at scale. Lithological attributes and characteristics present in available geoscience data can be correlated with related remote sensing data using a machine learning model, which can then infer lithological attributes and characteristics for regions where remote sensing data is available, but geoscience data is not.

A method is described for seismic imaging including receiving a pre-stack seismic dataset and an earth model at one or more computer processors; performing least-squares reverse time migration of the pre-stack seismic dataset using the earth model to create a digital seismic image, wherein the least-squares reverse time migration includes wave-equation forward modeling based on an asymptotic expression for reflection in a subsurface Kirchhoff integral; and generating a display of the digital seismic image on a graphical user interface.

Methods, apparatuses, and mediums related to a full-wavefield angle gather generation for high-contrast inter-thin bed modeling for reservoir characterizations of a survey region are provided. A method may include a well logging tool having one or more sonic generators and one or more well log data recording sensors in a wellbore. Sound waves may be generated using the one or more sonic generators in order to generate reflections in the survey region. Well log data, based on the reflections, may be received using well log data recording sensors, and the well log data may be transmitted to at least one memory. A method may perform, using a computer system, a full-wavefield angle gather generation. A method may generate, by a computer system, the high-contrast inter thin-bed models based on the full-wavefield angle gather.

Disclosed are methods, systems, and computer-readable medium to perform operations including: receiving an input dataset that represents partial spatial information of an area of interest; providing the input dataset to a spatial context generator, wherein the spatial context generator comprises a machine learning model trained to generate, based on the partial spatial information, contextual spatial information for the area of interest; and using the spatial context generator to generate, based on the partial spatial information, at least one output dataset associated with the area of interest, where each output dataset comprises simulated contextual spatial information for the area of interest.

Determination of an impulse response at a subsurface image level can include extrapolation of an up-going pressure wavefield to a subsurface image level, extrapolation of a down-going velocity wavefield to the subsurface image level, and determination of the impulse response at the subsurface image level from a hypothetical seismic source by spectral division of the extrapolated up-going pressure wavefield by the extrapolated down-going velocity wavefield.

A computer-implemented method for analyzing geophysical data is disclosed. Interpretation of geophysical data, such as seismic data, can be performed in multiple stages, such as at an information extraction stage and an information analysis stage. Typically, the information analysis stage is performed by geologists or interpreters, which may be laborious and inconsistent. The disclosed method includes using an information extractor that extracts information indicative of geo-features in a subsurface and an inference engine that analyzes the information indicative of geo-features in a subsurface to generate an output, with the information extractor and the inference engine being integrated and acting in combination. For example, the information extractor may generate summaries of the geo-features or answers to questions. In this way, the information extractor and the inference engine in combination may act synergistically, such as in the context of reasoning, natural language processing, and the outputs generated.

Some systems and methods for determining depths of subterranean formation layer tops while drilling through the subterranean formation include a drill bit, a drill rig, a microphone, a depth sensor, and a processor. While drilling the through the subterranean formation, the processor receives a measured sound from the microphone and a measured drill bit depth from the depth sensor, normalizes the measured sound across all measured drill bit depths, determines frequency information of the normalized sound for each depth of the plurality of depths, determines frequency spectrums of the normalized sound for one or more depths of the plurality of depths, transforms the frequency spectrums into a depth spectrum, and determines the depths of subterranean formation layer tops based on the depth spectrum.

A global objective function is initialized to an initial value. A particular model simulation process is executed using prepared input data. A mismatch value is computed by using a local function to compare an output of the particular model simulation process to corresponding input data for the particular model simulation process. Model objects associated with the particular model simulation process are sent to another model simulation process. An optimization process is executed to predict new values for input data to reduce the computed mismatch value.

A method and apparatus for performing geological reasoning, A method includes: obtaining subsurface data for a subsurface region; obtaining a knowledge model; extracting a structured representation from the subsurface data using the knowledge model; and performing geological reasoning with a graph network based on the knowledge model and the structured representation. A method includes performing geological reasoning with a knowledge model that includes a set of geoscience rules or a geoscience ontology. A method includes performing geological reasoning with a structured representation that includes a graph. A method includes performing geological reasoning by one or more of the following: question answering; decision making; assigning ranking; and assessing probability.

A subsurface representation may define simulated subsurface configuration of a simulated subsurface region. The simulated subsurface region may include simulated wells, and the simulated subsurface configuration may define simulated correlation between the simulated wells. Subsurface configuration of wells may be compared with the simulated subsurface configuration to generate similarity maps for the wells. Simulated wells may be matched to the wells based on the similarity maps and the arrangement of the wells. Correlation between the wells may be determined based on the simulated correlation between the matched simulated wells.