A Hybrid Analytical–Genetic Programming Framework for Modelling Cross-Shore Beach Profiles

Description

1. Overview This database contains all datasets generated and used in the study focused on the identification, classification, and modeling of nearshore sandbar morphology using in situ bathymetric profiles, statistical analysis, genetic programming, and satellite-derived morphometric indicators. The database supports the full methodological workflow, including data pre-processing, dimensionality reduction, clustering, machine learning modeling, remote sensing processing, and validation.

Steps to reproduce

2. Data Structure The database is organized into the following main directories: 2.1 Raw Data (/Duck_original/) Contains the original, code for acquisition with metadata from Duck, NC: Bathymetric survey profiles (cross-shore profiles) Original geographic coordinates Raw depth measurements Formats: .txt, .mat, .csv. Content: No filtering, interpolation, or transformations applied. 2.2 Pre-Processed Data (/02_preprocessed_data/) Profiles after standardization and preparation: Removal of the emerged beach and swash zone Horizontal alignment of profiles Interpolation to a regular 1 m grid Truncation to 600 m offshore extension Removal of missing or invalid values Output: Uniform bathymetric profile matrices Corresponding acquisition dates 2.3 PCA and Clustering (/03_pca_clustering/) Dimensionality reduction and classification results: K-means cluster assignments Centroid profiles Validation Metrics: Silhouette Index PERMANOVA test results 2.4 Morphologycal Parameters (/04_morphologycal_parameters/) Computed parameters used as predictors in the modeling stage: Bar crest position (𝑥𝑐) Bar trough position (xf) Depth at bar crest (ℎ𝑐) Depth at bar trough (ℎ𝑓) Depth at shoreline reference (ℎ𝑡​) Bar height index (ℎ𝑥) 2.5 Genetic Programming Models (/05_genetic_programming/) Results of the symbolic regression modeling: Simplified analytical equations Model performance metrics (R², MAE, RMSE) Sensitivity analysis outputs

Institutions

• Universidade Federal de Santa Catarina

  SC, Florianopolis

Categories

Licence