Modeling Tsunami Deposit Grain Size Trends Using Particle Tracking: Implications For Flow Inversions

Tsunamis are some of natures most catastrophic events, displacing great amounts of water, sediment, and people alike. Sediment deposits left behind by these rare but extreme events are often the only records of their occurrence. Deposit characteristics - especially grain size distributions (GSDs) - can, in principle, be used to put bounds on flow magnitudes, wave heights and velocities. However, there are many uncertainties in how the flows transport and sort sediments because direct observations are difficult. We developed a model tracking the motion of particles in the water column to gain a better understanding of grain advection, dispersion, and the GSDs deposited by tsunamis. In our model, we explicitly impose and adjust variables including turbulent fluctuations, mean velocity and velocity profiles, bedload, resuspension probabilities, and source particle GSDs and positions, and measure the effects on particle trajectories and deposition. We compare the particle-tracking model results to controlled laboratory experiments to better interpret experimental trends. We also use the particle-tracking model outputs of deposit grain size distributions as inputs for an advection-settling inversion model for flow depth and velocity, in order to systematically evaluate how different variables (turbulent dispersion, resuspension, etc.) influence paleoflow predictions. Preliminary results suggest that a single deposit grain size percentile (e.g., D50) does not provide an accurate prediction of flow depth and velocity along the deposit, but that the best-fit deposit percentiles change with transport distance. Different vertical profiles for the mean velocity appear to have relatively little impact on deposit particle distributions. However, inverse model results are fairly sensitive to downstream transport by bedload and resuspension. By using our model to systematically vary parameters that are difficult to control using physical experiments (e.g., turbulence, resuspension probabilities), we hope to better unlock interpretations of paleotsunami and storm surge hydrodynamics from their deposits.