Geophysics
Reza Ghanati; Zahra Tafaghod Khabaz
Abstract
Inversion modeling of Vertical Electrical Sounding (VES) data is formulated as a nonlinear problem and solved using multiple depth layers of fixed boundary and a few depth layers of the variable boundary. The model parameters in Smooth inversion are only the resistivity values. However, in the block ...
Read More
Inversion modeling of Vertical Electrical Sounding (VES) data is formulated as a nonlinear problem and solved using multiple depth layers of fixed boundary and a few depth layers of the variable boundary. The model parameters in Smooth inversion are only the resistivity values. However, in the block inversion, the thickness of each layer is also added to the model parameters. Due to the non-linearity of the inverse problem, the determination of an appropriate initial model is very significant. A technique has been adopted to estimate a proper starting model for the block inversion strategy. Thus, from the solution of the smooth inversion, the second derivative of cumulative resistivity is calculated, then using the difference of turning points, the initial values of the model parameters are determined. The proposed algorithms are first tested on data derived from two artificial models and then implemented on two real data sets. The numerical experiments demonstrate that the smooth inversion of the geo-electrical sounding data due to less dependency on the initial model confronts fewer challenges proportional to the block inversion strategy. However, in the case of block inversion, an appropriate choice of the initial model makes it possible to determine the boundaries of the layers with more certainty.