Investigation of accretion profile caused by cross-shore sediment movement with physical model

In this study, by taking in the consideration of wave height and period, bed slope and sediment diameter, cross-shore sediment movement was studied in a physical model and various berm geometric parameters and of the resultant accretion type profile were determined. Also sorting of sediment on accretion profile were examined. Regression analyses were made using of physical model data, then dimensional and non-dimensional equations were found for each parameter. Classical regression analyses and multivariate adaptive regression splines techniques were used to generate the equations. In addition, artificial bee colony (ABC), teaching–learning based optimization (TLBO) and JAYA algorithms were applied to regression functions to obtain dimensional and non-dimensional equations. The experimental results were also evaluated by using artificial neural networks (ANN) and the related berm geometric parameters were determined. Dimensional and non-dimensional ANN models were obtained by using back propagation, ABC, TLBO and JAYA algorithms. Results obtained from the tests were compared with results from the developed models and the equations given in the literature. It was concluded that the results of the equations and neural networks were better than those of cited in literature. Also, foreshore of accretion zone, crest of accretion zone, ridge of accretion zone, shoreline and toe of accretion zone were determined as sections in accretion profile than sorting of sediment in these sections were determined. In consequence of this study, it was clearly understood that finer material was accumulated on the ridge of accretion profiles, but coarser grains deposited on the front, shoreline and toe sections. There was not an assessment for crest of accretion zone.