Long-Term, Ensemble, Data-Assimilated Coastal Change Modeling

We present an ensemble Kalman filter shoreline change model to predict long-term coastal evolution due to waves, sea-level rise, and other natural and anthropogenic processes responsible for sediment transport. The model presented here utilizes ensemble simulations to improve both reliability (via data assimilation) and uncertainty quantification. We assess uncertainty across a variety of physical processes and spatiotemporal scales using a straightforward technique based on the spread of the (deterministic) model ensemble with perturbed parameters and forcing. Coastal change projections exhibit significant differences when simulated with and without ensemble wind-wave conditions. Many long-term coastal change projections rely on a single realization of the future wave climate, often derived from atmospheric conditions simulated by a global climate model. Yet, a single realization approach does not account for the stochastic nature of future wave conditions across a variety of temporal scales (e.g., daily, weekly, seasonally, and interannually). Here, by applying ensemble time series of wave forcing conditions, produced by a computationally efficient statistical downscaling method, we demonstrate a sizable increase in model uncertainty compared with the unrealistic case of model projections based on a single realization (e.g., a single time series) of the wave forcing. We support assessments of model uncertainty with analytical results derived from idealized examples. In the present study, we apply the developed ensemble modeling approach to a well-monitored beach in Tairua, New Zealand. However, the model is generally applicable to a variety of coastal settings around the world.