A holistic approach to inversion of time-domain airborne EM
4
Citation
4
Reference
10
Related Paper
Citation Trend
Abstract:
A "holistic" method for simultaneously estimating conductivity and calibration models from 1-D inversion of time-domain airborne electromagnetic (AEM) data is proposed. The work extends the concept of holistic inversion that been successfully applied to frequency-domain AEM data. The entire multi-component airborne dataset and available independent conductivity and interface-depth data are simultaneously inverted. A spline-based conductivity model covering the complete survey area is estimated. Unmonitored elements of the system geometry are included as unknown parameters of the calibration model and are solved for in the inversion.Conventional 1-D inversion methods invert each airborne sample in isolation from other samples. However, by simultaneously considering all of the available information together in a holistic inversion formulation, we are able to exploit the inter-component and spatial coherency characteristics of the airborne data. The formulation ensures that the conductivity and calibration models are optimal with respect to the airborne data and prior information.Keywords:
Inverse transform sampling
Abstract As an important branch of geophysical inversion, non‐linear inversion method has played a unique role in geophysical inversion. In recent years, more researchers are laying emphasis on non‐linear joint inversion, but no substantial progresses have been made at present. This paper tries to realize a two‐dimensional non‐linear inversion based on the high density resistivity method by using a joint calculation of both Back Propagation neural network (BP neural network) method and ant colony algorithm which is called Ant colony optimized Back Propagation (ABP) method. After a comparison of results concluded from two models, calculation with BP and ABP methods could achieve better results than with conventional method. Furthermore, the ABP method is superior to the BP neural network method for it could reduce number of iterations, save computation time and obtain a higher calculation accuracy.
Inverse transform sampling
Backpropagation
Ant colony
Cite
Citations (13)
Inverse transform sampling
Synthetic data
Seismic velocity
Cite
Citations (3)
Three dimensional gravity inversion is an effective way to extract subsurface density distribution from gravity data. Different from the conventional geophysics-based inversions, machine-learning-based inversion is a data-driven method mapping the observed data to a 3D model. We have developed a new machine-learning-based inversion method by establishing a decomposition network (DecNet). Unlike existing machine-learning-based inversion methods, the proposed DecNet method is a mapping from 2D to 2D, which requires less training time and memory space. Instead of learning the density information of each grid point, this network learns the boundary position, vertical center, thickness, and density distribution by 2D-to-2D mapping and reconstructs the 3D model by using these predicted parameters. Furthermore, by using the highly accurate boundary information learned from this network as supplement information, the DecNet method is optimized into a DecNet B method. By comparing the least-squares inversion and U-Net inversion on synthetic and real survey data, the DecNet and DecNet B methods have shown the advantage in dealing with inverse problems for targets with boundaries.
Inverse transform sampling
Prior information
Cite
Citations (14)
Abstract Gravity inversion is a typical geophysical inversion method that obtains the underground density distribution by analyzing the gravity anomaly. Normally, it can be divided into geophysics-based and deep learning based inversion. The 3D geophysics-based inversion is a time- and memory-consuming method, so 3D inversion is not routinely implemented in practical data interpretation. Here, we propose a deep learning method to transfer the 3D inversion problem to a multiple layers 2D mapping problem by decomposing the 3D target into four 2D images, including the horizontal location, vertical center, thickness and density distribution. This method is denoted as “decomposition network”. By implementing synthetic experiments with regular and complex models, and comparing with the 3D U-Net inversion, the proposed network has proved can reconstruct underground targets with high accuracy and high efficiency.
Inverse transform sampling
Cite
Citations (0)
The three‐dimensional gravity inversion based on the properties model is very popular in recent years. The tendency of intelligent algorithm applied for inversion is developing consistently. Yao et al. (2003) and Jian Zhang et al. (2004) discussed many of the problems for gravity inversion with a genetic algorithm. However, the inversion will be very difficult with the increasing parameters in the very huge searching space. Therefore, used the relatively few parameters to describe the inversion target is very important for improving the inversion efficiency. The purpose of our study was to reduce the parameters of three‐dimensional gravity inversion and to lighten the image quality of inversion result. The results show that the new method can expedite the inversion processing and get a better geological interpretation than tradition method.
Cite
Citations (1)
Inverse transform sampling
Hydrocarbon exploration
Cite
Citations (1)