We use a landscape evolution model to infer the effect of Late Pleistocene climate change on the incision-aggradation behaviour of the Rhine-Meuse fluvial system. We model the routing of runoff and sediment in the catchment in order to predict grainsize trends and the incision and aggradation behaviour in the downstream reach, where we compare it to the sequence of events and grainsize characteristics inferred from borehole corings. This sequence starts with an important incision taking place around the MIS 3 to MIS 2 climatic transition. During the coldest part of MIS 2, a coarse-grained sedimentary unit is deposited that shows an upward increase in the sand/gravel ratio. The model experiments do not predict an incision at the MIS 3 to MIS 2 transition. Therefore, the incision should be attributed to other causes, most likely effects of glacio-isostatic uplift. However, a relative upward increase in sand content of the sediments is predicted by the model. This increase is the result of the difference in transport rates between sand and gravel. Starting from a homogeneous pre-existing (MIS 3) deposit, the gravel content in the active layer increases because the sand is removed quickly and transported further downstream, whereas the gravel travels slowly and piles up with gravel originating from immediately upstream, resulting in a net accumulation. At a later stage, sand originating from much further upstream progrades fan-like over the gravelly deposits. According to the record, during the early Late Glacial warming part of MIS 2 (Bolling-Allerod interstadial), neither incision nor aggradation has taken place. This is in accordance with modelling results which show that, despite the reduction of sediment input due to re-vegetation of hillslopes, sufficient sediment remains available for fluvial transport in the channel network itself. It takes several thousands of years before effects of sediment depletion in the catchment are noted downstream. That is why we argue that the inferred incision at the late Late Glacial (the start of the Younger Dryas) in our downstream study area might reflect depletion effects related to the preceding early Late Glacial conditions. In general, our modelling results show that terraces along one large fluvial system are diachronic features. in particular, terrace surfaces are older upstream compared to downstream. In addition, complex responses to climate change are likely to occur in a large fluvial system like the Rhine-Meuse, and correlation of morphological features in the fluvial record to specific short term palaeo-climatic events, for example Dansgaard-Oeschger events could be risky without consideration of catchment (size) characteristics and associated response times. (C) 2009 Elsevier B.V. All rights reserved.

Climate change scenarios based on integrations of the Hadley Centre regional climate model HadRM2 are used to determine the change in the flow regime of the river Rhine by the end of the 21st century. Two scenarios are formulated: Scenario 1 accounting for the temperature increase (4.8degreesC on average over the basin) and changes in the mean precipitation, and Scenario 2 accounting additionally for changes in the temperature variance and an increase in the relative variability of precipitation. These scenarios are used as input into the RhineFlow hydrological model, a distributed water balance model of the Rhine basin that simulates river flow, soil moisture, snow pack and groundwater storage with a 10 d time step. Both scenarios result in higher mean discharges of the Rhine in winter (approx. + 30%), but lower mean discharges in summer (approx. -30%), particularly in August (approx. -50%). RhineFlow simulations also indicate that the variability of the 10 d discharges increases significantly, even if the variability of the climatic inputs remains unchanged. The annual maximum discharge increases in magnitude throughout the Rhine and tends to occur more frequently in winter, thus suggesting an increasing risk of winter floods. This is especially pronounced in Scenario 2. At the Netherlands-German border, the magnitude of the 20 yr maximum discharge event increases by 14% in Scenario 1 and by 29% in Scenario 2; the present-day 20 yr event tends to reappear every 5 yr in Scenario 1 and every 3 yr in Scenario 2. The frequency of occurrence of low and very low flows increases, in both scenarios alike.