A method is described for analysis of subsurface data including the use of physics-based modeling and experimental design that allows calculation of probabilities of physical subsurface properties. The method may include calculations of key controlling parameters. The method may include using multiple dimension scaling. The method may be executed by a computer system.

A method for estimating uncertainties in determining hypocenters of seismic events occurring in subsurface formations according to one aspect includes determining estimates of event locations by choosing local peaks in summed amplitude of seismic energy detected by an array of sensors disposed above an area of the subsurface to be evaluated. For each peak, the following is performed: recomputing the summed amplitude response for a selected set of points of comprising small perturbations in time and space from the estimated event locations; computing second derivatives of log likelihood function from the stacked responses at the estimated location and the perturbed locations; assembling the second derivatives into a Fisher information matrix; computing an inverse of the Fisher information matrix; determining variances of estimated parameters from the elements from the diagonal of the inverted matrix; and computing standard deviations of the estimated parameters by calculating a square root of the variances.

A method for interpreting seismic data includes receiving seismic data that represents a subterranean volume, and generating inline probability values and crossline probability values using a first machine learning technique. The first machine learning technique is trained to identify one or more vertical fault lines in a seismic volume based on the seismic data. The method includes generating a merged data set by combining the inline probability values and the crossline probability values, training a second machine learning technique based on a subset of labeled horizontal planes from the merged data set, the second machine learning technique trained to identify horizontal fault lines from the seismic volume, and generating a representation of the seismic volume based on the second machine learning technique, the representation including an indication of a three-dimensional fault structure within the seismic volume.

Systems and methods for updating a reservoir model are disclosed. In one embodiment, a method of updating a computer model includes receiving actual well data from a plurality of wells, and accessing model well data from the computer model for a plurality of modeled wells, wherein the plurality of modeled wells correspond to the plurality of wells. The method further includes comparing, by a computing device, the actual well data to the model well data according to a grid model vertical mismatch metric. When the grid model vertical mismatch metric is satisfied, the method includes globally updating the computer model. When the grid model vertical mismatch metric is not satisfied, the method includes comparing the plurality of wells of the actual well data to a cluster metric, when the cluster metric is satisfied, locally updating the computer model, and when the cluster metric is not satisfied, globally updating the computer model.

A method is described for assessing subsurface structure uncertainty based on at least one subsurface horizon. The method calculates seismic continuity attributes to determine a mappability of the subsurface horizon(s); determines horizontal uncertainty for each fault in vertical uncertainty for each horizon; generates probabilistic scenarios for a subsurface geometry for at least one conceptual model; and generates a map of geological model uncertainty based on the probabilistic scenarios. In some embodiments, the probabilistic scenarios are stochastic simulations. In some embodiments, generating a map of geological model uncertainty is based on information entropy. The method may be executed by a computer system.

A method is described for assessing subsurface structure uncertainty based on at least one subsurface horizon. The method calculates seismic continuity attributes to determine a mappability of the subsurface horizon(s); determines horizontal uncertainty for each fault in vertical uncertainty for each horizon; generates probabilistic scenarios for a subsurface geometry for at least one conceptual model; and generates a map of geological model uncertainty based on the probabilistic scenarios. In some embodiments, the probabilistic scenarios are stochastic simulations. In some embodiments, generating a map of geological model uncertainty is based on information entropy. The method may be executed by a computer system.

Aspects of the present disclosure relate to confidence volumes for earth modeling using machine learning. A method includes receiving detected data, wherein the detected data includes formation attributes relating to one or more depth points along or near a wellbore. The method further includes providing inputs to a plurality of machine learning models based on the detected data. The method further includes receiving output values from the plurality of machine learning models based on the inputs. The method further includes determining a measure of variance among the output values. The method further includes generating a confidence indicator related to the output values based on the measure of variance.

Unsegmented continuous subsurface data may be analyzed using one or more moving windows to characterize a subsurface region. Unsegmented continuous subsurface data may be scanned using the moving window(s). Probabilities that portions of the subsurface region include a subsurface feature may be determined based on analysis of the portions of the unsegmented continuous subsurface data within the moving window(s).

A system includes a processor and a memory. The memory includes instructions that are executable by the processor to access a plurality of seismic images of a subterranean formation in a first geological area. The instructions are also executable to generate a plurality of fault estimates from each of the plurality of seismic images. Further, the instructions are executable to generate a processed seismic image of the first geological area by normalizing and merging the plurality of seismic images and the plurality of fault estimates. Additionally, the instructions are executable to generate a statistical fault uncertainty volume of the first geological area using the processed seismic image. Furthermore, the instructions are executable to control a drilling operation in the first geological area using the statistical fault uncertainty volume of the first geological area.

The invention is a method applicable to oil and gas exploration and exploitation for automatically detecting geological objects belonging to a given type of geological object in a seismic image, on a basis of a priori probabilities of belonging to a type of geological object assigned to each of samples of the image to be interpreted. The image is transformed into seismic attributes applied beforehand, followed by a classification method. For each of the classes, an a posteriori probability of belonging to a type of geological object is determined for each of the samples of the class according to the a priori probabilities, of the class, of belonging, and according to a parameter α describing a confidence in the a priori probabilities of belonging. Based on the class of the sample, the determined a posteriori probability of belonging to a type of geological object is assigned for the samples of the class. The geological objects belonging to the type of geological object are detected based on determined of the a posteriori probabilities of belonging to the type of geological object for each of the samples of the image to be interpreted.