Evolutionary polynomial regression model for the prediction of coastal dynamics

The effective protection of the coastal ecosystem requires a detailed knowledge of the morphological evolution of the coastal environment. Several probabilistic models have been developed in the last decades to implement a reliable statistical forecasting of coastline dynamics. In this work, the non-linear Evolutionary Polynomial Regression (EPR) model has been used for the first time to evaluate the short-term dynamics of the shoreline from a set of measured shoreline positions in previous years. A comparison of the mean known shoreline positions with those predicted by the model, together with their confidence and prediction intervals, can be used to assess the reliability of the estimation by the EPR model.