1.5D internal multiple prediction on physical modeling data

Pan (Penny) Pan, Kristopher A. Innanen, Joe Wong

Multiple attenuation is a key aspect of seismic data processing, with the completeness of multiple removal often significantly affecting final image results. In this paper, we analyze 1.5D internal multiple prediction on physical modeling data simulating a 2D marine seismic survey designed to generate significant internal multiples. We describe a processing flow appropriate for preparation of the data for input into multiple prediction. Then we examine a 1.5D (i.e., pre-stack data over a layered geology) implementation of the inverse scattering series internal multiple prediction. The results show good agreement of predictions compared against synthetic data and physical modeling data. We discuss the selection of the integral limit parameter ? and the influence of free-surface multiples. We also demonstrate that the beginning and ending integration points of frequencies, wavenumbers, and pseudo-depths in the code can be optimally chosen to reduce computational burden.