Quantitative interpretation of multiple self-potential anomaly sources by a global optimization approach

Abstract

Multiple self-potential (SP) anomalies are analyzed by using a Genetic-Price Algorithm (GPA), which has been recently introduced for the inversion of SP data. The proposed approach is tested on multiple synthetic anomalies, which are modeled by horizontal cylinders. First, a forward modeling is used to analyze the resolution of such anomalies by varying all model parameters. Then, GPA is applied to invert synthetic multiple SP anomalies. The numerical analyses show that the proposed approach is able to fully characterize the anomaly sources by providing the correct values of the model parameters as well as the number of sources, even if Gaussian random noise is added to the synthetic data. Furthermore, to show the computational efficiency of GPA, the results of a comparative analysis with the Very Fast Simulated Annealing algorithm are given. The validity of the GPA approach is confirmed by its application to three examples of self-potential field data from mineral exploration and groundwater investigations, which are presented and discussed in relation to other inversion approaches. Finally, the quantitative interpretation of multiple anomalies along a SP profile crossing the Mt. Somma-Vesuvius volcano caldera (southern Italy) is provided.