Most Energetic Traveltime Depth Migration with Neural Networks

There is an extensive body of literature demonstrating that Kirchhoff Depth Migration using first-arrival travel times often produces suboptimal migrated images in regions with complex geology. To address this limitation, researchers since the 1990s have explored the use of the most energetic travel times for Kirchhoff migration, instead of relying solely on first arrivals (see Figure 1a and 1b for a comparison). For instance, it has been shown that this approach yields more accurate and reliable subsalt images and CRP gathers. However, the primary drawback is the significant increase in computation time, as calculating the most energetic arrivals often requires solving the wave equation, which is more computationally intensive than solving the eikonal equation used for first arrivals. Focusing on the 2D case, this study aims at reducing the computational burden of generating max-energy-arrival times by leveraging deep learning techniques.