Data Assimilation and Bathymetry Inversion at Tidal and Navigational Inlets

Bathymetric surveys can be difficult at tidal and navigational inlets, yet knowledge of the changing bathymetry is important to many applications (e.g. navigational planning, dredging). Estimation of bathymetry using only remote sensing observations would enable a more continuous view of the evolving bathymetry at an inlet. Herein, we report on progress towards this goal at two sites with drastically different dominant dynamics that were the focus of the research program DARLA: New River Inlet, NC, and Mouth of the Columbia River, OR/WA.

We use an ensemble-based data assimilation methodology employing a modeling system based on coupled circulation and wave models (specifically, the COAWST modeling system). The models are two-way coupled, and therefore account for the influence of the tidal currents on the waves, but also the influence of wave forcing on the circulation. The use of such a coupled modeling system also enables us to utilize observations of different geophysical variables. In particular, we utilize surface current maps obtained using an airborne Synthetic Aperture Radar system, and wavelength information (for several frequency bands) obtained using a tower-mounted X-band radar system. We contrast the two sites by discussing data availability and coverage, information content in each observation type (with respect to bathymetry estimation), and influence of the dominant dynamics on the bathymetry estimates. Finally, we explore the efficacy of the method with only short-dwell observations of currents and waves (hence, not relying on long-dwell observations over multiple tidal cycles) and report on the resulting bathymetry estimates.