Systems and methods for estimating reservoir productivity as a function of position in a subsurface volume of interest are disclosed. Exemplary implementations may: obtain an initial depth uncertainty model; obtain training depth uncertainty parameter values from the non-transient storage medium; obtain corresponding training depth uncertainty values; generate a trained depth uncertainty model by training the initial depth uncertainty model using the training depth uncertainty parameter values and the corresponding training depth uncertainty values; and store the trained depth uncertainty model.

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.

A system and method for determining an uncertainty of a distance to bed boundary (DTBB) inversion of a geologic formation. The system or method includes receiving logging data from a borehole tool, performing a first DTBB inversion using the logging data to calculate first DTBB solutions, adding quantified noise to the logging data to produce an adjusted signal, performing a second DTBB inversion using the adjusted signal to calculate second DTBB solutions, comparing the first DTBB solutions to the second DTBB solutions to determine an uncertainty of the first DTBB solutions based on a relationship of the quantified noise and the difference between the first DTBB solutions and the second DTBB solutions.

A method and system of machine learning-augmented geophysical inversion includes obtaining measured data; obtaining prior subsurface data; (a) partially training a data autoencoder with the measured data to learn a fraction of data space representations and generate a data space encoder; (b) partially training a model autoencoder with the prior subsurface data to learn a fraction of model space representations and generate a model space decoder; (c) forming an augmented forward model with the model space decoder, the data space encoder, and a physics-based forward model; (d) solving an inversion problem with the augmented forward model to generate an inversion solution; and iteratively repeating (a)-(d) until convergence of the inversion solution, wherein, for each iteration: partially training the data and model autoencoders starts with learned weights from an immediately-previous iteration; and solving the inversion problem starts with super parameters from the previous iteration.

A method and apparatus for generating a fluid saturation model for a subsurface region. One example method generally includes obtaining a model of the subsurface region; for each of a plurality of fluid types: flooding the subsurface region model with the fluid type to generate a flood model; and running a trial petrophysical inversion with the flood model to generate a trial petrophysical model; identifying potential fluid contact regions in the trial petrophysical models; partitioning the subsurface region model at the identified potential fluid contact regions; and constructing the fluid saturation model from the partitioned subsurface region model.

The subject technology relates to synthetic aperture to image leaks and sound sources. Other methods and systems are also disclosed. The subject technology includes drilling a wellbore penetrating a subterranean formation. The subject technology includes logging the wellbore using the stationary acoustic sensor and the moving acoustic sensor of the logging tool to obtain logged measurements, and obtaining an actual acoustic signal associated with a leak source in the wellbore using logged measurement data. The subject technology also includes determining a synthetic acoustic signal indicating an estimated leak source in the wellbore, and determining a correlation between the synthetic acoustic signal and the actual acoustic signal. The subject technology also includes generating a probability map from the determined correlation, in which the probability map indicates a likelihood of the leak source being located at a given location in the wellbore based on the probability map.

In some embodiments, an apparatus and a system, as well as a method and an article, may operate to generate a parent population, wherein each member of the parent population includes a set of model parameters describing a layer model of the geological formation; to execute a perturbation algorithm to generate subsequent child populations, from the parent population, until a termination criterion is met; to provide a plurality of solutions based on at least one member of the parent population and on at least one member of each child population; and to control a drilling operation based on a revised layer model that has been generated based on a selected one of the plurality of solutions. Additional apparatus, systems, and methods are disclosed.

A method and system for modeling a subsurface region include applying a trained machine learning network to an initial petrophysical parameter estimate to predict a geologic prior model; and performing a petrophysical inversion with the geologic prior model, geophysical data, and geophysical parameters to generate a rock type probability model and an updated petrophysical parameter estimate. Embodiments include managing hydrocarbons with the rock type probability model. Embodiments include checking for convergence of the updated petrophysical parameter estimate; and iteratively: applying the trained machine learning network to the updated petrophysical parameter estimate of a preceding iteration to predict an updated rock type probability model and another geologic prior model; performing a petrophysical inversion with the updated geologic prior model, geophysical seismic data, and geophysical elastic parameters to generate another rock type probability model and another updated petrophysical parameter estimate; and checking for convergence of the updated petrophysical parameter estimate.

Uncertainty in microseismic monitoring sensor data can be reduced. A computing device can receive information about at least one sensor that is monitoring a subterranean formation, including a location, after a fracturing fluid is introduced into the formation. The computing device can also receive information about a microseismic event and determine a seismic ray bath between a location of the event and the at least one sensor, and an uncertainty value of the location based on information about the formation and the information about the event. The computing device can determine a total uncertainty value associated with the locations of a plurality of microseismic events, including the microseismic event. The computing device can determine a solution to an objective function based on the total uncertainty value and a number of sensors. The computing device can determine a new location of the at least one sensor based on the solution.

A method of estimating lateral positioning uncertainties of a seismic image is performed at a computer system. The computer system receives a velocity model, the velocity model including a plurality of base velocity values used for generating the seismic image, each base velocity value having a low limit and a high limit. The computer system derives a plurality of lateral velocity gradient uncertainties from the velocity model and generates multiple lateral velocity gradient profiles, each lateral velocity gradient profile including a random realization of the plurality of lateral velocity gradient uncertainties. The computer system calculates perturbation raypaths originating from a surface point of the seismic image based on the velocity model and the multiple lateral velocity gradient profiles and estimates a lateral positioning uncertainty for a target location at a predefined depth of the seismic image based on lateral distributions of the perturbation raypaths at the predefined depth.