Work

A hybrid, genetic programming and physically-based predictor of dune geometry

Year: 2024
Type: article
Abstract: Subaqueous sand dunes are found in many natural environments and pose significant operational challenges. However, classic dune predictors found in the literature fail at predicting equilibrium dune dimensions. In this study, we first investigated the potential of using genetic programming to derive predictive equations of dune wavelength and height. The predictors outperformed existing relationships, yet the equations remain complex due to the intricate physics governing dune evolution. We carried out a global sensitivity analysis to evaluate the most influential parameters of the GP predictors. Finally, we proposed a set of robust predictors, for equilibrium dune heights and wavelengths, relying on basic environmental parameters.
Language: English
Open Access status: hybrid
Estimated APC paid: $3,300
OpenAlex ID: W4404291920
FWCI: 0.611
Cites: 114
Cited by: 3
Related to: 10