Internal Multiple Attenuation by Iterative Construction of Virtual Events

To obtain an accurate internal multiple estimation without any subsurface information, a new data-driven method is proposed. This novel method is based on the modified concept of virtual events. With the rigid theory deduction, we give the modified version of virtual events construction and its practical iterative calculation method. With the new definition of virtual event, the effectiveness and verification of this new method will be shown in several synthetic data. The differences between the method and traditional data-driven multiple attenuation algorithms will also be shown in these examples. The results demonstrate that the internal multiples can be effectively predicted and subtracted from original dataset by the new method and that the new method has advantages over traditional virtual events method in accuracy of internal multiple prediction.