A comparative analysis of SP data inversion by spectral, tomographic and global optimization approaches

Self-Potential (SP) fields are natural fields that originate from various forcing mechanisms related to electrical, hydraulic, chemical and thermal gradients. Due to the complexity of the source mechanisms, inversion of SP data is not easy and motivates the development of suitable techniques depending on application field, which ranges from engineering and geotechnical investigations to geothermal and mineral explorations. In this work, quantitative interpretations of self-potential data are given when SP anomaly sources can be modelled by simple polarized bodies whose parameters have to be determined. In particular, a comparative analysis is performed for the solutions of three different methods based on highresolution spectral analysis, tomographic approach and global optimization, respectively. The efficiency of each technique has been tested by finding depth, polarization angle and shape factor of the anomaly source on synthetic data generated by simple geometrical structures (like sphere, horizontal and vertical cylinder and inclined sheet) and on field examples. The study shows limits and potentialities of the investigated methods and suggests hybrid algorithms as suitable tools for an accurate and full characterization of the anomaly source.