Improvement in Subsurface Resistivity Image, Acquired from 2D Geoelectrical Tomography Survey, by Using Combined Weighted Inversion of Two Different Electrode Arrays Data.

By focusing on the sensitivity patterns and characteristics of electrode arrays, it can be realized that some arrays e.g. Pole-Dipole and Dipole-Dipole have high lateral resolving power. Some other arrays e.g. Wenner and Schlumberger have good vertical resolving power. In most electrical tomography studies, it has been experienced that by using one array, inversion of data cannot produce an optimum image of subsurface. When the Geoelectrical survey is done by using two electrode arrays, which each of them belongs to one of two different groups of arrays that has been mentioned above, an optimum image of the subsurface can be produced by using combined weighted inversion method. This optimum result is achieved because of employing merits of two electrode arrays in a single time. At the present research, Wenner and Pole-Dipole arrays have been used, and their data were inverted in combined routine by using smoothness-constrained inversion algorithm. Because, electrical potential sensitivity of two arrays and also their influences on inversion procedure are not equal, weighted combined inversion method was performed in order to equalize the portion of influence of each individual array on inversion process. Considering the fact that, Jacobin matrix is same with sensitivity matrix that relates model parameter variations to the data variations, therefore Jacobian matrix was employed to equalize the portion of influence of two arrays in inversion process. Also Finite Difference method was used to perform forward modeling which is needed along the inversion iterative stages. This method has good compromise between precision and time of calculations. A MATLAB code which accomplishes forward modeling by Finite Difference method was employed for using through inversion procedure and also it was used to produce synthetic data for every predefined model. Another computer program was developed in MATLAB for inversion algorithm calculations. Synthetic data of Wenner and Pole-Dipole arrays were inverted in single and in weighted combined manner. For all states, it was found out that results of weighted combined inversion are more near to reality of the defined structure, than single inversion of each array data. Anyhow suggested method seems to be more reliable in study of complicated subsurface structures.