Magnetotelluric sounding(MT)is an important geophysical method which has been widely used in mineral resource exploration,deep structure exploration,and other fields.Currently,MT is further developing towards refinement,and resistivity and magnetic permeability are two important physical parameters that affect the response of MT,and the phenomenon of resistive anisotropy objectively exists.However,mature MT 3D data interpretation techniques usually only consider a single parameter of resistivity and have poor scalability.Therefore,this article considers both resistive anisotropy and magnetic permeability parameters,and based on C language and MPI parallel programming model,implements an efficient large-scale multi-layer MT 3D parallel forward algorithm with parallel calculation between frequencies,parallel filling of impedance matrix,parallel addition of first type boundary conditions,and parallel solution of equation sets.On this basis,the influence of magnetic permeability on MT response was deeply studied.The experimental results show that the resistive anisotropy and the magnetic permeability make the MT response more complex.When the difference in resistivity between the anomalous body and the surrounding rock is small,larger magnetic permeability can even reverse the morphology of apparent resistivity and phase.This algorithm has a good scalability.For the test model;the 16 frequency serial computation takes 117.58 hours,while the 2048 process computation only takes 12.95 minutes,with an acceleration ratio of 547.24.The research results of this article contribute to further understanding the MT response characteristics of three-dimensional resistive and magnetic permeability anomalies,and provide efficient and fast forward modeling solutions for practical data interpretation.
维普
万方数据