Fluvial response to glacial-interglacial cycles - modelling the evolution of the Hochrhein using EROS

Repeated alpine glaciations during the Quaternary partly reached far into the foreland and caused profound landscape changes beyond glacial margins in northern Switzerland. Climate-driven glacier growth and decay and commensurate variations in water and sediment delivery caused river systems to aggrade and to incise, leading to the widespread occurrence of glacio-fluvial deposits (Deckenschotter) and associated terraces. Mapping and numerical dating of these depositional complexes increasingly offer insights into the spatial patterns and timing of Quaternary glaciations, and associated changes in (glacio-)fluvial dynamics. An improved understanding of how fluvial systems respond to glacial-interglacial cycles will help to assess the erosion potential around repository sites of nuclear waste over the next one million years. In this study, we contribute to close this research gap using numerical landscape evolution modelling (LEM).

We use EROS, a numerical landscape evolution model, which implements a particle-based approach to simulate water and sediment fluxes that interact with topography through erosive and depositional actions. Unlike LEMs based on the stream-power incision law, the method solves the 2D shallow water equations with both basal and lateral erosion and deposition, which allows for variations in width and lateral mobility of rivers; these variations induce changes in the transport capacity of the sediments that cause specific patterns of deposition and erosion to emerge. We adjusted the model so that it is capable to run over 1 Myrs, and imposed boundary conditions that - informed by estimates on longterm erosion rates – reflect rock uplift and plausible variations in water and sediment fluxes following a 100-kyrs glacial cycle. Our model relies on digital elevation models and sediment thickness data with 60 m spatial resolution and is applied to the Hochrhein river between Stein am Rhein and Basel and the Aare river downstream of the area, where Limmat and Reuss enter.

Our simulations show that the model reproduces widespread aggradation, reworking of the sediments by highly laterally mobile, braided river systems and incision during periods of increased runoff and low sediment availability.

Our model setup and parametrization features several uncertainties. For example, the capacity of rivers to laterally erode strongly determines the thickness and extent of depositional complexes lining the Hochrhein and Aare system. Also, our model is sensitive to temporally varying boundary conditions of water and sediment input about which precise estimates are lacking. Regardless, the more detailed and realistic representation of hydraulic and sediment transport processes by EROS compared to conventionally used landscape evolution models at this spatial and temporal scale provide the opportunity to test different hypotheses using numerical experiments and link the results to field evidence. Further sensitivity analyses and uncertainty quantification will enable us to use our model as simulation tool to hindcast and investigate the behaviour of fluvial system in response to different tectonic and climatic scenarios, thus helping to better understand potential spatial patterns of and sediment assemblages within widespread glacio-fluvial deposits.