The multiscale edge detection filter applied to airborne electromagnetic data interpretation: A case study at the Rio das Velhas Greenstone Belt, Quadrilátero Ferrífero, Brazil

• Published in: Journal of South American earth sciences Vol. 119; p. 103966
• Main Authors: Menezes, Paulo T.L., Barbosa, Valeria C.F., Pereira, Ronaldo M., Salomão, Marcelo
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-11-2022
• Subjects: Electromagnetic data; Greenstone belts; Mineral exploration; Non-magnetic shear zones mapping; Electromagnetic data; Mineral exploration; Greenstone belts; Non-magnetic shear zones mapping
• ISSN: 0895-9811; 1873-0647
• DOI: 10.1016/j.jsames.2022.103966

Potential-field edge detection filters are powerful fault detection tools in mineral exploration. The application of these filters to electromagnetic (EM) data is not as accepted as those associated with potential field data. But they may be advantageous when dealing with mineralizations associated with non-magnetic rocks. This paper introduces the multiscale edge detection filter to provide an enhanced image of fault zones when applied to apparent conductivity data. The Rio das Velhas Greenstone Belt, the studied area, hosts several gold mineralizations controlled by northwest–southeast shear zones subparallel to non-magnetic metasedimentary rocks that are typical candidates for the EM interpretation. Several northwest–southeast trending faults were highlighted in the studied area. We employed a data fusion interpretation strategy by correlating the detected EM faults with radiometric anomalies to classify high-priority spots for future exploration. We tagged four known gold deposits and new targets with similar characteristics following the proposed guidelines.
•We show that the multiscale filter (worming) can be used for electromagnetic (EM) data.•The EM worming screened non-magnetic mineralized fault zones in the Rio das Velhas Greenstone Belt.•Combining EM-worming and radiometric data, we selected sites for follow-up surveys.