Double-parameter inversion algorithm for RMT data

Abstract

The invention discloses a double-parameter inversion algorithm for RMT data, which comprises the following steps: S1, reading the positions of measuring points, the measuring frequency and the observation data of each measuring point, and constructing an N-dimensional observation data vector d; S2, assigning initial values to the two parameters of a model to be inverted; S3, calculating the forward response obtained by joint action of the two parameters according to the positions of the measuring points and the measuring frequency; S4, calculating the data fitting difference; S5, if the data fitting difference is greater than a target fitting difference, carrying out kth double-parameter inversion iteration, otherwise, ending inversion; and S6, adding 1 to the number of iterations, and returning to S3, wherein the kth double-parameter inversion iteration includes the sub-steps of calculating a sensitivity matrix of the forward response to the two model parameters, solving a double-parameter inversion iteration equation set to get the amount of correction of the two model parameters, and calculating new model parameters. The data fitting deviation can be reduced, and the accuracy ofinversion can be improved.

Description

technical field [0001] The invention relates to a high-frequency electromagnetic data inversion algorithm applied in the field of geophysics, in particular to the inversion of radio frequency magnetotelluric data. Background technique [0002] At present, the inversion of all kinds of electromagnetic data is aimed at seeking the distribution of subsurface medium resistivity that satisfies the poor fitting of the target, ignoring the influence of the difference in subsurface permittivity on the observed data. However, in practice, the observed data are jointly determined by the subsurface resistivity and permittivity, so there is a certain deviation in fitting the observed data with the traditional single-parameter resistivity inversion. Especially for Radio Magnetotelluric (RMT) data, the fluctuation field caused by the dielectric constant can account for more than 20% of the total field. In this case, only the inversion of resistivity parameters Fitting the objective funct...

AI Technical Summary

Problems solved by technology

[0002] At present, the inversion of various electromagnetic data is aimed at finding the distribution of subsurface medium resistivity that satisfies the poor fitting of the target, ignoring the influence of the difference in subsurface permittivity on the observed data

However, in practice, the observed data are jointly determined by the subsurface resistivity and permittivity, so the traditional single-parameter resistivity inversion is used to fit the observed data to a certain extent.

Especially for Radio Magnetotelluric (RMT) data, the fluctuation field caused by the dielectric constant can account for more than 20% of the total field. In this case, only the inversion of resistivity parameters Fitting the objective function will inevitably cause a certain deviation, which will reduce the accuracy of data interpretation

Images