Abstract:The basement relief of rift basin is a non-smooth interface, which cannot be delineated by traditional gravity inversion method. Besides, the gravity observing surface does not coincide with the upper interface of the sedimentary layers in most situations, which should be an up-and-down surface alone. Therefore, the basin model should be a dual interface model composed of an upper interface and its basement. We have developed a 2D gravity inversion method to estimate basement relief of rift basin based on a dual interface model, it can be directly used with gravity data observed on rough terrain. The density contrast between the sediments and the basement is assumed to be decreased monotonically with depth according to the hyperbolic law. The sedimentary layers can be divided into a set of 2D vertical, juxtaposed prisms, whose horizontal sizes are assumed to be known. The top surfaces of prisms coincide with the upper interface of the sedimentary layers. The depths of their bottom surfaces represent the depth of the basement and are the parameters to be calculated from gravity data. We have introduced the total variation (TV) function as constraint to the sedimentary basin model in order to estimate the basement relief caused by faults. The result will be non-smooth, which can well accord with the basement relief of the rift basin. We also use a few known depth points to reduce the multiplicity of the inversion. We build an objective function composed of data misfit function, known depth misfit function and TV function, then we use nonlinear conjugate gradients (NLCG) algorithm to minimize it. Tests conducted with synthetic data show that the inversion method can estimate basement relief of rift basin and also can estimate basement relief of downwarped basin by changing the regularization parameter. We apply the method to process the field data from Huizhou depression in Pearl River Mouth Basin and Yuncheng-linfen Rift Basin, respectively. The results clearly reflect the characteristics of the basement, which can provide important information for researching the structure of sedimentary basin and oil-gas exploration.

Recovering 3D Basement Relief Using Gravity Data Through Convolutional Neural Networks

3D Inversion of Gravity Anomaly – an Application of Technology of Apparent Density Imagery

3-D Gravity Inversion Based on Deep Convolution Neural Networks

Deep learning‐based inversion with discrete cosine transform discretization for two‐dimensional basement relief imaging of sedimentary basins from observed gravity anomalies

Deep Learning for Gravity and Magnetic Data Interpolation

Inversion of Gravity Anomaly Based on AlexNet Deep Neural Network

2D Gravity Inversion of Basement Relief of Rift Basin Based on A Dual Interface Model

3D gravity inversion based on deep learning

Joint Gravity and Magnetic Inversion Using CNNs' Deep Learning

3-D Gravity Inversion Based on Cross Dimensional Deep Neural Networks

Deep Learning for 3-D Inversion of Gravity Data

3D hybrid imaging based on gravity migration and regularized focusing inversion to predict the Poyang Basin interface

3D gravity inversion using cycle-consistent generative adversarial network

Recovering 3D Basin Basement Relief Using High‐Precision Magnetic Data Through Random Forest Regression Algorithm: A Case Study of Tianzhen‐Yanggao Sag in Datong Basin

A Deep Learning Gravity Inversion Method Based on a Self-Constrained Network and Its Application

High-resolution Seismic Impedance Inversion Integrating the Closed-Loop Convolutional Neural Network and Geostatistics: an Application to the Thin Interbedded Reservoir

Recovering Bathymetry of the Gulf of Guinea Using Altimetry-Derived Gravity Field Products Combined via Convolutional Neural Network

A Multitask Deep Learning for Simultaneous Denoising and Inversion of 3-D Gravity Data

Three-Dimensional Gravity Inversion Based on Attention Feature Fusion

Self-Supervised Deep Learning for 3D Gravity Inversion

Deep Subsurface Pseudo-Lithostratigraphic Modeling Based on Three-Dimensional Convolutional Neural Network (3D CNN) Using Inversed Geophysical Properties and Shallow Subsurface Geological Model