Forward seismic modelling, a tool for the prediction of offshore sedimentary architectures. Application to the Roussillon Basin's Pliocene land-see prism (Gulf of Lion, France).

The prediction of offshore sedimentary architectures is a first plan approach to the geological study of continental margins. While such work is commonly led using seismic surveys tied to well-logs, we are interested in land-sea objects for which there is no direct tie between seismic signal and lithology. More precisely, we look at shoreline clinoforms and more continental deposits of which size is below used seismic resolution, and their integration at the shelf-scale. The Roussillon Basin’s Pliocene infill satisfies these criteria. It belongs to a progradational land-sea prism about 100km-long, displaying essentially clinothems and defining the Gulf of Lion modern shelf. It is described with high quality conventional seismic profiles offshore, while outcrops and drill-cores are available onshore. However, there is no data at the transition between the two domains.

In order to predict the offshore sedimentary architecture, we establish classical equivalence hypotheses between seismic facies and expected associated sedimentary facies. This work is based on the seismic facies interpretation and on the lithologies known from outcrops and onshore drillings.

Nonetheless, without directly tied-in seismic such hypotheses rely essentially on interpretation. This, together with seismic data vertical resolution (~15m in thickness) and the upscaling from direct observations onshore, introduce uncertainties.

In order to produce more reliable sedimentary predictions, we test our hypotheses through forward seismic modelling using SeisRoX pro by NORSAR. We create small scale geological/impedance models based on onshore sedimentary observations coupled with well-logs petrophysical data (P-wave velocities). Then we simulate acoustic waves propagation through them and obtain theoretical seismic profiles that are subsequently compared to the seismic data.

This method, including a geophysical control, allows for the testing of various geological hypothesis at the outcrop-scale, and for a more objective subsurface description.

Among the results, we show that vertical velocity variations at a meter scale eventually get a specific seismic signature in terms of both geometry and amplitude on conventional seismic profiles. More generally, we illustrate different lithological models and their results, which allow for a high-resolution reconstruction of most parts of the Roussillon Basin’s Pliocene offshore prism.