Changing coastlines: data assimilation for morphodynamic prediction and predictability

Operational coastal flood forecasting is limited near-shore by lack of knowledge of evolving bathymetry. Obtaining a full bathymetric survey can be time-consuming and expensive, and is not generally possible just prior to a storm event. A more attractive option may be to use a coastal area morphodynamic model with data assimilation to maintain up-to-date model bathymetry in anticipation of such an event. Data assimilation provides a means to combine model predictions with observations to give more accurate forecasts. It also enables uncertainties in the forecasts to be calculated. Predictions of future bathymetry with specified uncertainty would be useful for coastal flood risk management purposes. In this paper we describe the application of data assimilation techniques to coastal area morphodynamic modeling. Two study sites are considered: the Dee estuary and Morecambe Bay.