CASCADE - A framework for modeling fluvial sediment connectivity and its application for drsigning low impact hydropower portfolios

Human activities, and especially dam constructions, have altered water and sediment fluxes in river basins with an unprecedented rate over the last decades. The resulting changes in river and delta processes pose a risk to river, floodplain, and coastal ecosystems and human livelihoods. While fluvial sediment transfers act on network-scales, impact assessments of dam sediment trapping (regarding downstream impacts and storage loss in reservoirs) and economic benefits focus commonly on single dam sites. Such local approaches omit the connected functioning of fluvial sediment transfers and the resulting network-scale trade-offs between dam economic benefits and cumulative sediment trapping of dams. Such analyses would require sediment data and models to evaluate future impacts of dams on network-scale sediment transport. However, neither sediment data nor numerical sediment transport models are commonly available on network scales. Within this context, the thesis sets out to answer the research questions of what would be optimal trade-offs between dam sediment trapping and economic hydropower benefits in large river systems? To answer this research question, the objectives of this thesis are 1) developing a network-scale sediment transport and connectivity model, 2) test its application for various, data scarce river basins with various initialization strategies and evaluate the robustness of results, and, 3) apply the model for quantifying cumulative dam sediment trapping and network-scale trade-offs between sediment trapping and hydroelectric production. First, we introduce a numerical modeling framework for connected sediment transfers on the scale of large river networks. The CASCADE (CAtchment Sediment Connectivity and DElivery) framework is a parsimonious statistical framework coupling network scale, grain-size specific sediment budgets with recent concepts of sediment connectivity. Analyzing sediment connectivity in two major Asian river basins (Da River, 50570 km\textsuperscript{2} and Se Kong, Se San, Sre Pok (3S) Rivers, 82500 km\textsuperscript{2}) quantified how sediment connectivity is a multi-scale, multi-domain property of river systems driven by the spatial distribution and properties of sediment supply and fluvial transport processes. Such information was so far available in qualitative terms, mostly, and only for smaller, well monitored rivers. CASCADE can be parameterized deterministically based on remotely sensed data using hydro-morphologic equilibrium considerations for large and data-scarce river systems. However, we also show how CASCADE allows to implement network-scale stochastic modeling for the 3S basin. Such stochastic modeling allows to dis-aggregate point observations of sediment flux and grain size into spatially distributed estimates of grain sizes and supply from many sediment sources to the river network via an inverse Monte Carlo Approach. This source information is used to model sediment flux and grain size composition in the entire river network . One key result is that sediment transport is spatially highly heterogeneous. This heterogeneity, which is derived from the CASCADE framework and validated with other lines of evidence, is observed for both case studies. This poses the opportunity to minimize dam sediment trapping by placing dams in parts of the networks where sediment transport is naturally low compared to the hydro-power potential. To analyse impacts of hydropower developments, we add a simplified model of reservoir hydraulics to CASCADE and analyze actual and hypothetical hydro-power development strategies for the 3S rivers. Developing all hydro-power dams will reduce total sediment flux from the 3S by more than 90 \%. Nearly 60 \% of this reduction are attributable to a single dam site. We then introduce a network scale analysis of dam portfolios (i.e., different combinations of the proposed dam sites). The aim of this analysis is to identify optimal trade-offs between dam hydro-power production, production costs, and sediment trapping. We find that, because of the spatial heterogeneity in sediment transfers, very similar hydropower production levels can be reached for very different levels of sediment trapping in function of how dams are placed within the river network. In the basin under study, up to 70 \% of the hydro-power potential could have been developed with a minor (20 \%) reduction in sediment flux. Additionally, using an empirical model for hydro-power production costs, we find that there is a strong synergy between providing cheap hydro-power and reducing dam sediment trapping on network scales. To conclude, CASCADE enables analysing and quantifying connected sediment transfers from a whole-network perspective. We provide evidence for how this novel information can be used to identify optimal dam portfolios that minimize trade-offs between hydropower production and sediment trapping in dams. Such a strategic planning of hydropower portfolios is advantageous from an environmental and economic perspective, and poses major potential for reducing impacts of future hydro-power developments on large river systems.