Neural network modeling of nearshore sandbar behavior

Abstract

The temporal evolution of nearshore sandbars (alongshore ridges of sand fringing coasts in water depths less than 10 m and of paramount importance for coastal safety) is commonly predicted using process-based models. These models are autoregressive and require offshore wave characteristics as input, properties that find their neural network equivalent in the NARX (Nonlinear Auto-Regressive model with eXogenous input) architecture. Earlier literature results suggest that the evolution of sandbars depends nonlinearly on the wave forcing and that the sandbar position at a specific moment contains ‘memory’, that is, time-series of sandbar positions show dependencies spanning relatively long time periods. Using observations of an outer sandbar collected daily for about 3.5 years at the double-barred Surfers Paradise, Gold Coast, Australia we find, however, little difference in performance between a NARX, an autoregressive multilayer perceptron (without long-term dependencies), and a linear NARX. It is uncertain whether these results generalize to the inner Gold Coast bar or to other field sites