Modeling water-layer-related multiples in ocean-bottom surveys and investigating associated elastic imaging artifacts caused by crosstalk

In a typical ocean bottom acquisition geometry, the source excites incident signals and generates abundant reflections from deep structures and multiples in the water layer. They are mixed and overlapped, and recorded by multicomponent ocean bottom receivers as seismic data. When imaging sub-seafloor structures, water-layer-related multiples can cause serious artifacts that deteriorate the depth images. In recent years, developing de-multiple techniques has always been a challenge. Although some methods were developed to suppress multiples, or separate multiples into different orders, in the reflection data, they are not always effective. Part of these challenges is that the reflection signals always coexist with multiples, both in real data and in numerical simulations. The above situation leads to a lack of an objective benchmark to effectively conduct quality control of these methods. In this research, by combining the implementation of the representation theorem with finite-difference (FD) simulation, we develop a method to synthesize multiples order by order. The synthetic data can be used for different purposes, e.g., investigating the properties of multiples and related artifacts, developing efficient de-multiple techniques, and examining the accuracy and reliability of existing de-multiple methods. As a straightforward application, we investigate imaging artifacts caused by crosstalk when inputting simulated water-layer-related multiples into an elastic reverse time migration (ERTM).