Jianxin Liu

Three-dimensional magnetotelluric regularized inversion based on smoothness-constrained model

How to get the rapid and stable inversion results and reconstruct the clear subsurface resistivity structures is a focus problem in current magnetotelluric inversion. A stable solution of an ill-posed inverse problem was obtained by the regularization methods in which some desired structures were imposed to stabilize the inverse problem. By the smoothness-constrained model and approximate sensitivity method, the stable subsurface resistivity structures were reconstructed. The synthetic examples show that the smoothness-constrained regularized inversion method is effective and can be reasonable to reconstruct three-dimensional subsurface resistivity structures.

Structure of the Upper Mantle and Transition Zone Beneath the South China Block Imaged by Finite Frequency Tomography

We applied the finite frequency tomography method to S wave data recorded by 350 broadband stations beneath the South China Block (SCB) and its surroundings from earthquakes occurring between July 2007 and July 2010, to better understand upper mantle deformation. Differential travel-times in the pair of stations with appropriate weighting for each station are used in the inversion. Our results are consistent with previous tomography that show a high velocity anomaly beneath the Sichuan basin and a high velocity anomaly in the transition zone beneath the Yangtze Craton. However, the resolution of mantle heterogeneity provides new insight into the tectonic framework of subduction of Burmese lithosphere in the west part of the study region and subduction of oceanic lithosphere in the east. In the subduction realm, west of 107°E, a significant fast S-wave anomaly is located on the southeast of Sichuan Basin. East of 107°E, and two narrow and discontinuous fast S-wave anomalies occur at a depth of 400–600 km beneath the middle of the South China block overlain by the pronounced low S-wave anomalies at a depth of 100 and 400 km. If the fast anomalies located in the mantle transition zone represent stagnant slabs, their fragmented nature may suggest that they could be produced by different episodes of subduction beneath western Pacific island and the above slow velocity anomaly may associated with the back-arc regions of ongoing subduction. In addition, tomography also reveals an anomalously high S-wave velocity continental root extends eastward to a depth 400 km beneath the eastern Sichuan Basin. This anomaly may be related to eastern extrusion of Indian lithosphere associated with the collision of India and Eurasia. Moreover, our results also show large slow anomalies beneath the Red River fault region connected to deeper anomalies beneath the South China Fold Belt and South China Sea. All these observations are consistent with the scenario that the South China block has been built by both of subduction of Paleo-pacific plate and eastward subduction of Burma microplate.

A decorrelation-based hybrid global search algorithm for inversion of 1D magnetotelluric data

An efficient algorithm is presented in this paper to improve the perturbation efficiency of the adaptive downhill simplex simulated annealing (ASSA) method for magnetotelluric inversion with oblique correlated misfit valleys. The correlated model space is rotated to a less dependent space which is defined by the eigenvectors of the parameter covariance updated progressively by the inversion itself. The downhill simplex step and rotation work together to decorrelate the correlated model space. The application to two synthetic cases and real data indicates that ASSA in the rotated space generally has a better convergence and efficient behaviour than ASSA without rotation. In the rotated space, the high rejection rates, which occurred in the unrotated space, are avoided. At low temperature, the estimated covariance can be used to approximate the global covariance. For all cases, ASSA in the rotated space gives better inversion results.

3-Dimensional Modelling for Direct Shear Test of Structure Plane in Rock Mass

Direct shear test is an important test in geotechnical engineering to obtain the basic parameters of the rock mass. In order to study the three dimensional characteristic of structure plane during the direct shear test, the numerical simulation software FLAC3D is used to build the calculation model for the structure plane under the compound stress of normal direction and shear direction. The deformation and mechanical response of the structure plane are analyzed, which show that, (1) Relationship between shear strength and normal stress meets in the Mohr-Coulomb linear way, which is the same as the laboratory test and theoretical analysis; (2) the distribution of stress on the structure plane increases from the side of loading to the other side; and with the increase of normal stress, while the distribution of maximum shear stress does not vary greatly. The analysis results can give some guidance for the real practice; (3) the result from the numerical modeling method is close to that from the laboratory test, which confirms the validity of the numerical method. Keywords-3-Dimensional Modelling; structure plane; direct shear test; numerical simulation; rock mass

A Hybrid Optimization Method Based on Genetic Algorithm for Magnetotelluric Inverse Problem

Genetic algorithm, one of the new methods for global non-linear optimization problem, has been applied in magnetotelluric data analysis. In this paper, the magnetotelluric inverse problem was studied by a hybrid genetic algorithm, which was based on the combination of simplex method and genetic algorithm. The standard genetic algorithm has poor local search ability, large amounts of calculation, and adaptability to large space. On the other hand, simplex method based on local linearization is usually lost in local minimum values. So a new method was put forward through a promoted simplex operator embedded into genetic algorithm and the strategy was adopted to keep the group of best individuals. The new algorithm has not only the global of genetic algorithm, but also the fast convergence of the simplex algorithm. The inversion results of synthetic magnetotelluric data shows that the algorithm possesses advantages of expediting convergence, avoiding earliness and improving precision.

Dynamic imaging of metallic contamination plume based on self-potential data

Abstract A dynamic imaging method for monitoring self-potential data was proposed. Based on the Darcys law and Archies formulas, a dynamic model was built as a state model to simulate the transportation of metallic ions in porous medium, and the Nernst equation was used to calculate the redox potential of metallic ions for observation modeling. Then, the state model and observation model form an extended Kalman filter cycle to perform dynamic imaging. The noise added synthetic data imaging test shows that the extended Kalman filter can effectively fuse the model evolution and observed self-potential data. The further sandbox monitoring experiment also demonstrates that the self-potential can be used to monitor the activities of metallic ions and exactly retrieve the dynamic process of metallic contamination.

Development of multi-channel observation and inversion for IP electrical sounding method

In order to improve the exploration effect of deep non-ferrous mineral resources, multi-channel observation methods for induced polarization (IP) electrical sounding data and their inversion imaging technology are studied. First of all, four multi-channel observation methods are developed based on conventional IP electrical method, namely three-electrode and four-electrode arrays of unilateral and bilateral current transmitting. Then the maximum smoothness constrained inversion method of the least squares sense for IP electrical sounding data is proposed, and the inversion software is programmed. Finally, the simulation and inversion results of geo-electrical model for the proposed observation methods are analyzed. And the comparison results show that three-electrode array of bilateral current transmitting gives the best result, but the intensity in field work is larger than others; unilateral three-electrode and four-electrode arrays give the better results. Taking detection results and convenience of field exploration work into consideration, these two methods are more suitable for practical application; bilateral observation method of four-electrode array is not suitable for the detection of the steep ore bodies.

Damped Gauss-Newton Optimization Algorithm for Two-Dimensional Magnetotelluric Regularization Inversion

The two-dimensional magnetotelluric inverse problem is ill-posed and the inverse results are unstable and non-unique. It means that different geo-electrical model could fit the observed data with the same accuracy. A stable solution of the ill-posed inverse problem can be obtained by utilizing the regularization methods in the objective function. Solving large scale linear equation of inverse problem, the damped Gauss-Newton algorithm was adopted, which can improve local convergence of Gauss-Newton method. Through the synthetic model simulation, the inversion results truly reflected the geo-electrical parameters of the model and accurately showed the depth and size of the abnormal body. On the one hand, inversion of TE mode was more sensitive for the low abnormal body and had poor resolution for the high abnormal body. On the other hand, inversion of TM mode had better resolution for the high abnormal body. So the damped Gauss-Newton algorithm can be used in two-dimensional magnetotelluric data analysis. Keywordsmagnetotelluric; two-dimensional inversion; damped Gauss-Newton; ill-posed; regularization

The Application of TSVD in MT Inversion with Improved MSE as Determination of the Regularization Level

The total mean-square error (MSE) of the estimated model, defined as the sum of the standard model variance and the bias variance, is used to define the truncation level of the singular-value decomposition to give a reasonable balance between model resolution and model variance. This balance is determined largely by the data and no further assumptions are necessary except that the bias terms are estimated sufficiently well. This principle has been tested on the 1D magnetotelluric inverse problem with special emphasis on high-frequency magnetotelluric (MT) data. Simulations clearly demonstrate that the method provides a good balance between resolution and variance. Starting from a homogeneous half-space, the best solution is sought for a fixed set of singular values. The model variance is estimated from the sum of the inverse eigenvalues squared, up to a certain threshold, and the bias variance is estimated from the model projections on the remaining eigenvectors. By varying the threshold, the minimum of the MSE is found for an increasing number of fixed singular values until the number of active singular values becomes greater than or equal to the estimated number. Keywordsinverse problem; TSVD; mean-square error; magnetotelluric model

Forward Modeling of Frequency Domain Controlled-Source Electromagnetic Using Finite Element Method

Frequency domain controlled-source electromagnetic (fCSEM) is an artificial source electromagnetic sounding method, which developed on the basis of magnetotelluric sounding. For showing the fCSEM forward modeling result, numerical simulation method must be adopted. In this paper, Base on Matlab as programming tool, we adopted the finite element method of quadratic interpolation in a rectangular element for solving the fCSEM forward problem and derived the corresponding formula for calculating. In order to simulate the infinity border and meet the demand computer memory, the nonuniform gird is designed with ensuring the accuracy. In programming, we only stored the non-zero elements of the finite element matrix for reducing the forward calculation time. By numerical simulation of homogeneous half-space model and 2-D model, we verified the correctness of the forward modeling algorithm. Keywordsforward modeling; frequency domain; controlledsource electromagnetic; finite element method