Optimized Crosshole Seismics for Near-Real-Time Ore Delineation in Active Mines

The resolution and accuracy of crosshole seismic tomography for mineral exploration are critically influenced by precise first-arrival picking and instrument positioning. We integrate neural networks for automated seismic arrival picking and a robust positioning-constrained inversion method to address these challenges. Utilizing high-frequency seismic data from medium range and deep boreholes, we demonstrate substantial improvements in tomographic image resolution and reliability. An automatic methodology for determination of P-wave and S-wave velocity fields and associated elastic parameters is effectively deployed, providing reliable lithological and mineralization insights. Recent advancements in automatic data processing and interpretation enable mining companies to significantly reduce exploration costs and improve resource estimation accuracy, facilitating more effective mine planning and development.