Propagation of hydro-geomorphic disturbances through continental-scale river basins

Hundreds of millions of people live close to, and depend upon, the world's large rivers for water, food, transport and the maintenance of a thriving ecosystem. However, these rivers are increasingly vulnerable to the effects of a wide range of natural and human-induced disturbances, including climate change, construction of large dams, river engineering works, deforestation, agricultural intensification, and mining activity. Over the past two decades, climate change and deforestation have impacted on the hydrology and sediment fluxes within the Amazon River Basin, and yet, the Amazon has remained one of the few large river systems that has been largely unaffected by dams. Nevertheless, because of extensive hydropower dam construction in Brazil, Bolivia, Peru and Ecuador now threatens the basin, with more 300 dams planned or under construction, this situation is changing rapidly. These dams are expected to trigger severe hydro-physical and ecological disturbances throughout the basin, including massive reductions in sediment and nutrient delivery to the lowland Amazon and its floodplains, substantial degradation of riverbeds and banks, significant changes in river water levels and flooding, and adverse impacts on river and floodplain ecosystems, on which the human population depends. There is a pressing need for action to assess and mitigate these impacts. However, our capacity to do this is severely restricted by an absence of quantitative models that can predict how environmental disturbances propagate through large rivers and floodplains, over continental distances, and decadal to centennial time periods. A key challenge in this respect is the need to develop models that are both physically-realistic and also computationally-efficient. The latter is critical for model application at the basin scale, and in order to derive large simulation ensembles that account for the substantial uncertainty in model parameters and environmental boundary conditions. We report here on the development and evaluation of such a model that operates at coarse spatial (10 km) and temporal (daily to annual) resolutions. Our new modelling approach simulates changes in river morphology (mean width, depth and slope), channel-belt topography (expressed as an elevation frequency distribution), and associated changes in flow conveyance, channel-floodplain connectivity and sediment delivery to downstream reaches. Model predictions are compared with, and evaluated against, simulations of river response to dam construction generated using a high-resolution physics-based modelling approach (with spatial and temporal resolutions of 50 m and <10 seconds). This comparison demonstrates that our new simplified model is able to reproduce the key trends in river evolution simulated by the physics-based model, and their dependence on the magnitude of the shift in hydrologic regime and sediment trapping efficiency for a range of environmental scenarios. Consequently, this new model may provide a suitable approach with which to evaluate the propagation of morphodynamic disturbances at the scale of very large basins, such as the Amazon.