Tunnel valleys have long fascinated the geoscientists by their scale and the consequent intensity of the sedimentary processes responsible for their formation. Tunnel valleys may be up to 180 km in length, up to4 km in width and 450 m in depth. The incisions are formed subglacially by overpressured meltwaters on the outer parts of continental­scale ice sheets. In addition to the
peculiar incision process, the filling of tunnel valleys in the southern and eastern North Sea, when imaged by 3D seismic data, show a peculiar infill. In these areas, the infill of the valleys is mainly composed of clinoforms prograding north in opposing direction to the former ice flows whose south­ward flowing meltwater excavated the valleys. The first hypothesis, by analogy with some small eskers introduced the concept of backfilling where the eroded sediments upstream is deposited directly below the ice margin, in a conveyor­belt fashion. The formation of the 'backsets' would have been enhanced by supercooling due to the pressure drop during the upward flow of the water from the deepest part of the valleys towards the ice margin, freezing and thus capturing the sediments on the adverse slope. Recently this model has been challenged by new observations on the architecture of the valleys and their infill sediments which appear to show many similarities in common with deltaic clinoforms observed in the pre­glacial succession of the southern North Sea.
The new model states that the incision and the filling of the valleys are separate in times and from distinct sedimentary processes. The valleys are mainly incised by overpressured subglacial meltwater with probably some abrasion as a minor erosive agent including bedrock control on the incision depths and morphologies. The infill is interpreted as proglacial for the newly observed south­dipping clinoforms and postglacial for the north­dipping clinoforms onlapping the latter. The north­dipping clinoforms are interpreted to be formed within a large deltaic system associated with the Rhine­Meuse river(s). The delta was probably infilling a large lake system containing overdeeps (the underfilled tunnel valleys). The presence of clinoforms 50­-80 m above the valley shoulders gives a fair idea of what could have been the general depth of the lake. However, the lake was certainly in the isostatic depression after partial or complete ice sheet retreat. The crust and mantle were not in equilibrium during that postglacial time, and the ice might have been occasionally present in the lacustrine basin so that the lake levels may have been very variable and difficult to tie to the present­day topographic configuration. This system of competition between one of the biggest river of Europe facing ice sheets and their proglacial depositional system generates a very intricate stratigraphy with multiple cross­cutting 'basins' in the form of valleys (c. 7 generations) which themselves contain up to 8 complete seismic sequences. Although the task to uild up a complete stratigraphic scheme is immense and a long run project, it would be unique for this period in the region. We intend to solve part of the problem by numerically reconstructing the local landscape with the ice sheet and its isostatic depression. This allows to represent a type-sequence helping the understanding of the Rhine-­Meuse migration and the position of the lacustrine systems which so far has remained elusive.