The EPIGRAM project concerns the dynamics of the French Atlantic marginal and coastal areas (Bay of Biscay and Channel). It is indeed noticeable that most of the researches on the dynamics of the ocean have concentrated on the deep ocean, and its low frequency (climate) variability. The coastal area has only driven back our attention in this field recently, thanks to the development of coastal operational oceanography, and the need for a better monitoring of this area for economical, health or military purposes. One of the main scientific lag is associated with observations at sea: there do not exist 'heavy' campaigns at sea covering this area, and most of the scientific data were collected 15 years ago or more. The project is therefore based on: The definition, realisation and exploitation of campaigns at sea, performed by IFREMER, CNRS/INSU and SHOM. The construction of realistic numerical models of the area, and their comparison with observations at sea, on the basis of process studies. The processes selected for this project are high frequency to seasonal variability at the most. We will concentrate on hydrodynamical studies and the main scientific goal is to improve our comprehension of the main dynamical processes of the continental shelf and margins in the 'Manche' (Channel) and 'Golfe de Gascogne' (Bay of Biscay) and the ability of the numerical models to represent them. The project gathers 14 oceanographic public laboratories with about 50 people. The expected outcomes of the project are: The realisation of four important campaigns at sea and the collection of data for all selected processes. The scientific analysis of the collected data (furniture of diagnostics for all selected processes). Validated realistic numerical models of the area (whose results will be compared to -and limits will be assessed from comparison with- observations at sea). An improved understanding of the major physical processes of the area. In summary, the present project is proposed for a four years period. It concentrates on the physics of the Bay of Biscay and the Channel, in particular on the continental shelf and margin dynamics. It is based on the definition and exploitation of campaigns at sea and numerical modelling studies. The general goal being to improve our understanding and modelling capacity of selected physical processes. These results will be of primary importance for the oceanographic scientific community in general and for nowcast/forecast operational systems. Finally the chosen approach (built on the analysis of oceanic processes) will also facilitate the extension of the results gathered on the present project to other regions.