Robust interpretation of single and multiple self-potential anomalies via flower pollination algorithm

Abstract

Self-potential (SP) method is suitable for identifying positions of fractures, ore bodies, leakage, corrosion of metal, etc. The identification of SP anomalies can be implemented through inversion process, which not only serves to determine the best model parameters but also to estimate their uncertainties. The model parameters and their uncertainty can be determined from the posterior distribution model (PDM) of the SP anomaly inversion. To find the required PDM, in this paper, generalized likelihood uncertainty estimation (GLUE) thresholding is proposed as a joint approach together with flower pollination algorithm (FPA). This overall method is then tested on several synthetics and fields (i.e., metallic drum, Weiss, Sawoo, and LUSI anomalies) SP data containing single and multiples of SP sources. The results demonstrate that the proposed algorithm is robust for solving quantitative interpretations of SP data. Moreover, this method does not require prior assumptions over the shape of the anomaly source.