Multi-parameter Controlled Automatically Picking And Variable Smoothing For Tomography With Fast 3D Beam Prestack Depth Migration

Automatically picking on 3D volumes and fast convergence to variable wavelength are two major challenges in grid-based reflection tomography. This paper discusses a multi-parameter controlled automatically picking method that, when integrated with a fast beam migration, provides reliable picks of primary events while avoiding coherent noises such as multiples. To speed up the convergence of the solution, we use a 3D Gaussian filter as a smoothing operator that enables variable smoothing lengths for different parts of the model. Combining the efficiency and flexibility of the beam migration with the two proposed developments, we provide a robust and economical tool for velocity model building. Field data tests show uplift in image quality when using an updated velocity model after one pass of the iterative solution.