Enhancing Seismic Image Quality through an Automatic Refraction Static Correction: A Machine Learning Application in Web Based Seismic Processing

Summary Onshore seismic survey in Indonesia often poses static problem due to highly heterogeneous near surface velocity structure. We implemented an automatic refraction static workflow with machine learning algorithm to perform first break picking. The subsequent step includes multi layer modeling, ray tracing, and tomo-static calculation. Application of the method to the East Java onshore 2D dataset has proven the effectiveness of the workflow. Significant improvement on the data quality has been achieved. Enhancement of event continuity as well as increasing signal to noise ratio have been successfully attained. The resulting shallow velocity-depth profile may also be utilized in the depth imaging shallow velocity model to obtain better imaging quality.