What is it about?
It is about computing an initial resistivity model from the 1D Schlumberger VES data. Few decades ago, and in some places in this generation, the partial curve matching results from the VES data used to be a standard interpretation for the 1D Schlumberger VES data. But with the availability of powerful nonlinear least-squares inversion schemes today such results could only be an approximation. However, this approximation is a useful initial model for the sophisticated nonlinear least-squares inversion schemes we have today. In the past, the partial curve matching was done manually; now I present a MATLAB code to automate the process. In summary, I present an automatic way of producing an initial resistivity model for a subsequent inversion procedure.
Featured Image
Why is it important?
Gradient based inversion usually require an initial model to begin the process. Initial model farther from the local minimum may not yield reliable results and sometimes the scheme may fail because of the poor choice of the initial model. But calculating the initial model from the data provides a kind of educated guess that will be useful. It saves time for making random guesses.
Perspectives
Most of the time, an initial model for nonlinear least squares inversion is guessed or retrieved from some expensive sources (e.g. well logs) which may not be available. However, since the data is always available it thus saves time and energy to get the initial model from it. I expect that this MATLAB code would be a useful research tool that will inspire similar products (automatically producing initial models) for more complicated and higher dimension geophysical problems (2D and 3D).
Jide Ogunbo
Read the Original
This page is a summary of: MATLAB code for data-driven initial model of 1D Schlumberger sounding curve, Geophysics, March 2018, Society of Exploration Geophysicists,
DOI: 10.1190/geo2016-0631.1.
You can read the full text:
Contributors
The following have contributed to this page
