The most important geophysical problems are inverse gravimetry and magnetometry problems. Among them are the structural gravimetry and magnetometry problems of finding interfaces between layers with different densitie...
详细信息
ISBN:
(纸本)9786197105100
The most important geophysical problems are inverse gravimetry and magnetometry problems. Among them are the structural gravimetry and magnetometry problems of finding interfaces between layers with different densities or magnetizations using known gravitational or magnetic data [1], [2], [3]. The gravimetry and magnetometry problems are described by nonlinear integral Fredholm equations of the first kind;they are ill-posed problems. After the discretization of integral operators, the problems are reduced to systems of nonlinear equations with dense matrices. The real gravity and magnetic measurements are carried out over a large area producing large-scale grids. Processing of gravity and magnetic data is time consuming and requires a lot of memory. In this paper, for solving the structural inverse magnetometry problem in a multilayer medium, efficient stable parallel algorithms based on iteratively regularized gradient methods with variable weight factors are proposed. The algorithms were implemented numerically with using new computing technologies on the parallel computing system Uran at the Institute of Mathematics and Mechanics of the UB RAS. The structural magnetometry problem with "quasi-model" data was solved.
暂无评论