Avulsion in action: Reconstruction and modelling sedimentation pace and upstream flood water levels following a Medieval tidal-river diversion catastrophe (Biesbosch, The Netherlands, 1421-1750 AD)

Deltaic land inundated by storm surges may reform by sedimentation from natural or human-induced river diversions. This is a well-known trigger mechanism for creation of new channels in coastal plains and deltas, which may develop into main channels and lead to abandonment of older (avulsion), particularly in the downstream parts of deltas that host tidal rivers. These new channels develop as part of deltaic splay complexes that heal initial diversion scars and fill up flooded basins at a certain pace. We study a case with excellent historical and geological data of a diversion of the river Rhine following catastrophic inundations (1421-1424 AD) into medieval reclaimed land. Numerical modelling of deltaic splay and channel development is combined with reconstructions from historical maps and geological data. This yields detailed insight in pacing of splay sedimentation and changing hydrodynamics in the channel upstream of the diversion in the two centuries following the inundation. The equivalent of the full sand budget of the river Rhine was effectively trapped in the developing splay. The tidal-avulsion splay evolution on aspects is similar to that of fluvial crevassing into flood basins documented for settings lacking 'downstream' tidal control. The typical small-scale delta-lobe avulsion cycles: mouth bar formation, backward sedimentation, upstream avulsion, channel progradation and mouth bar formation are reproduced in the splay-modelling. The pacing of splay development, however, is relatively fast due to the presence of tides and the water depth in the receiving basin. The diversion had a strong upstream impact, in particular on water levels in the feeding river channel at stages of peak flow. For two centuries levels were significantly raised, because bifurcation-imposed reduced transport capacity and associated sedimentation at the diversion site increased hydraulic roughness and hampered flow. These findings have implications regarding flood mitigation for presently-planned lower-delta engineered diversions. Furthermore, they elucidate the differences between upstream fluvial avulsion and downstream tidal-river avulsion that are important to recognise if we want to understand how deltaic distributary-networks are maintained. (C) 2009 Elsevier B.V. All rights reserved.