Multiples from the sea bed or salt boundary are usually strong, and will interfere and generate artifacts on migration images. In order to attenuate these multiples, one approach is to first predict and then subtract them. We use a model-based prediction method that requires a velocity model which contains the strong multiple generating interfaces, such as the sea bed or salt boundary. This method is also a two-way waveequation approach in the sense that it solves the full wave equation by finite-difference approximation in the prediction. After prediction the modeled multiples are subtracted from the data with a matching filter. This strategy is tested on a synthetic data set. Numerical results show that a cleaner migration image is obtained after the multiples are attenuated with this strategy.
展开▼