Evolutionary modeling of longshore sediment transport

Longshore sediment transport is one of the main causes of coastal erosion, being generated for natural reasons or because of man-made structures and activities. Therefore, its assessment is very important in coastal engineering design and analysis (e.g., for design of breakwaters at harbor entrances, navigation channels, etc.). However, such transport process is very complex due to magnitude and direction varying seasonally, daily or even hourly, because of the variability in waves reaching the shore. Previous studies show that numerous formulas and models for computing the sediment transport by waves and currents have been proposed, ranging from quasi-steady formulas based on the traction or energetics approach to complex numerical models involving higher-order turbulence closure schemes. This paper investigates the longshore sediment transport rate as related to the most important physical parameters influencing the phenomenon, using a data-driven approach based on the Evolutionary Polynomial Regression (EPR) modeling paradigm. The study employs an extensive literature database including field measurements. The results are very interesting, returning sound formulations containing all the most important physical parameters that are known to influence the analyzed process.