Alternative 3D Tomography Methods and Their Applications to Identify Seismic Structure Around the Hydrothermal Gas Field

Abstract

The use of 3D tomography methods helps to the better understanding of complex structures peculiar to elements of a large-scale system, for instance, to buildings, engineering constructions, hydroelectric power plants, mines ets. Seismic tomography is important tool that the oil and gas industry applies to map the sub-surface geological structures. The LSQR algorithm by Paige and Saunders is the well-known conjugate gradient solver for the inversion tomography problem. Nevertheless, previous studies have shown that the algorithm has limitation in reconstruction of synthetic models of the large-size structure. Here two alternatives are given to resolve the complex structure: a new algorithm of the method of coordinate descent and the modification of Gaussian elimination. Both of these methods aim at constructing of the solution, which might be free from influence of determined types of error in the data. A structure with certain similar properties has been revealed for the layer from 0 to 5 km of depth by applying both methods to seismic data collected in the area around the hydrothermal gas field However, for the deeper layer the modification of Gaussian elimination produces the structure that is more realistic compared the method of coordinate descent. The reason might be in more accurate determination of the location of those seismic sources that are close to a surface. We conclude that simultaneous application of different inversion methods is useful in the presence of unclear level of errors in the data. Tomography images presented here for the area of natural concentration of gases can be applied for understanding of conditions to keep underground gases in a safe area.