Refined Techniques for Data Processing and Two-Dimensional Inversion in Magnetotelluric (Ⅲ)——Using the Impressing Method to Construct Starting Model of 2D Magnetotelluric Inversion
详细信息   
摘要


     In the Mainstream MagnetotelluricMT 2D inversion currently based on the linear optimization technology, the construction of starting model has a great influence on the inversion result. How to construct reasonable starting model for the 2D MT inversion is still an urgent problem. A starting model of uniform half space or from 1D inversion result may not easily obtain reliable inversion effect. Theory and long-time MT inversion experience shows that despite the difference of 2D magnetotelluric inversion results based on different starting models, the results all are closer to the true model than the starting model and obtain macro-outline characteristics ofthe true model. But in some details, they may be biased. Based on the understanding of such a phenomenon, and after the repeated exploration in theory and practice, we propose the Impressing method for the starting model building in MT 2D inversion, of which the basic idea is to determine the next 2D inversions starting model by weighting models between uniform half space and the existing inversion result. The starting model built by the Impressing method contains the real information of true model and meanwhile, ensures the uniformity of deep electric structure which meets magnetotellurie 2D forward bottom boundary condition requirements. Based on the Impressing method, this paper further put forward the idea of multi-iterative inversion. Through the 2D inversions of several theoretical models and field measurement MT data, we verify the good effect of the above-mentioned methods on suppressing the non- uniqueness of inversion results resulting from starting model construction and so improving the reliability of MT 2D inversion results. This may also provide a new idea for constructing the starting model of other geophysical inversion.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700