CONSTRAINING EXTREME EVENT MAGNITUDES USING DEPOSIT GRAIN SIZE DISTRIBUTIONS: UNCERTAINTIES AND OPPORTUNITIES

Climate describes not only mean environmental conditions but also extreme events that can be recorded imperfectly in stratigraphy. In principle, mechanistic relations for predicting fluvial sediment transport rates can be used to quantitatively infer the magnitudes of flow events preserved in flood, storm surge, and tsunami deposits. Using (1) laboratory flume experiments and (2) a numerical particle-tracking model, I explore how deposit grain size distributions (GSDs) can be inverted to calculate flow depths and velocities. The flume experiments were initially designed to represent scaled and idealized tsunamis or storm surges, although the underlying concepts can also be applied to floodplain sediment advected away from channels during floods. The experiments demonstrate how source GSDs, transport distances, flow depths and velocities, and turbulent dispersion influence deposit GSDs. Second, I present a new numerical forward model for creating synthetic deposits, in which grain trajectories are tracked as a function of imposed mean flow, turbulent intensity, and source GSD. Both the experimental and modeled data are used to rigorously evaluate predictions of event magnitudes using existing advection-settling and transport-capacity based models. Equally important is quantifying inversion uncertainty, which comes from many sources including poorly known boundary conditions. For tsunami and storm surge applications, boundary conditions such as source GSDs and transport distances from source to deposit are usually known or assumed. However, in the absence of constraints on these boundary conditions, I explore how well or poorly relative event magnitudes could be inferred from a vertical succession of event bed GSDs. Finally, two additional issues can inherently limit our ability to infer distributions of event magnitudes: morphodynamic feedbacks (which change the relationship between discharge and transport capacity), and also the implicit assumption that sediment transport is an equilibrium process.