Validation And Analysis Of A 1-D Variational Assimilation Scheme For Bathymetry Inversion

Results are presented from a variational data assimilation and nearshore bathymetry inversion scheme based on depth and wave-phase averaged surfzone hydrodynamics for a narrow-banded wave spectrum. 1-D Cross shore bathymetry is inverted based on 17-min-averaged significant wave height and longshore current observations made during the Duck94 field experiment. An inversion scenario is considered in which inversion is carried out separately for each 17-min-averaged set of observations. Results are used to quantify the inversion scheme's accuracy depending on time dependent wave and bathymetry conditions. The model generally produced a rootmean-square error of 0.46 m between modelled and actual bathymetry when both wave height observations and alongshore current observations were assimilated. The RMS error reduced to 0.44 m when only wave height observations were assimilated and increased to 0.69 m when only alongshore current observations were assimilated, suggesting that wave height is a more informative variable for bathymetry inversion. The model performed consistently in inverting bathymetry and thereby tracking morphologic change in varied wave conditions, including during large storm events.