Computation for Inversion of Two and Three-Dimensional Magnetotelluric Data

PI发展了一种适用于二维和三维大地电磁数据的有效反演方法，称为快速松弛反演(RRI)。该方法是迭代的，用前一次迭代的值来近似电场和磁场的水平导数。然后，每个位置下的电导率摄动反问题就变得相互解耦，并与一维问题密切相关。该方法计算速度非常快，并且计算时间只随数据倒置站点的数量而线性增长。真实数据集的2-D数据可以在高端技术工作站上进行反演。这笔赠款支持进一步的工作，以解决某些计算困难，研究大型二维问题，并将该方法扩展到现实的三维问题。

The PI has developed an efficient inversion method applicable to both two and three-dimensional magnetotelluric data called the Rapid Relaxation Inverse (RRI). The method is iterative and approximates horizontal derivatives of the electric and magnetic fields with their values from the previous iteration. Then the inverse problems for the conductivity perturbation beneath each site become uncoupled from each other and closely related to the 1-D problem. The methoda is very fast, and computation time grows only linearly with the number of sites of data inverted. 2-D data for realistic data sets can be inverted on high end technical workstations. This grant supports further work to resolve certain computational difficulties, to work on large 2-D problems and extend the method to realistic 3-D problems.