VEMI: a flexible interface for 3D tomographic inversion of time- and frequency-domain electrical data in EIDORS
G. De Donno and E. Cardarelli
Journal name: Near Surface Geophysics
Issue: Vol 15, No 1, February 2017 pp. 43 - 58
Info: Article, PDF ( 1.89Mb )
Price: € 30
The aim of this work is to present a MATLAB interface called Versatile interface for Electrical data Modelling and Inversion included in the EIDORS software. The interface is able to invert 2D and 3D electrical data acquired in the time or the frequency domain, both for cylindrical and prismatic geometries. Therefore, it can be a flexible tool for inversion of both real- and complex-valued resistivity tomography data acquired in the laboratory and in the field. The forward solver has been proved to be stable and accurate through a comparison with an analytical solution, whereas the inverse solution, achieved through a Gauss–Newton routine with an optimised damping inner loop, can be performed with the help of other useful tools incorporating a priori information in the inversion process. The reliability of the Versatile interface for Electrical data Modelling and Inversion has been tested through a 3D laboratory example, where both time- and frequency-domain data and two 3D field example with time-domain data were acquired. The 3D laboratory example, simulating a shallow aquifer contaminated by a chlorinated solvent (hydrofluoroether), demonstrated the reliability of the Versatile interface for Electrical data Modelling and Inversion to detect the contaminant pathway within the physical model. Hydrofluoroether is clearly visible on phase and chargeability models, where the highest phase values are located underneath the spilling point, even though it remains undistinguishable in the resistivity and amplitude ones. Through the combined analysis of the inverted chargeability and phase models, we can reduce the degree of uncertainty in the interpretation of geophysical models. These results were also validated through comparison with the respective synthetic models simulated in a previous paper by the same authors. Real-world tests have been performed on a closed landfill where few a priori information are available about the original design and on an industrial site contaminated by chlorinated solvents. In the former case, we reconstruct a three-layer configuration (covering, waste and bottom liner), where the effective layering inferred from the resistivity model is confirmed by the chargeability one. In the latter case, we detect the chlorinated solvents within the deeper aquifer through a combined analysis of the resistivity and chargeability models, where the very high resistivity values are associated with high chargeability.