Shoreline predictive model using artificial intelligence for the homogeneous beach of the Western Coast of Ghana
Coastal areas are preferred home to about 40% of global population than any other ecosystems. Many organisms, including endangered species, live along sandy beaches. Nonetheless, such beaches are prone to erosion and flooding owing to natural and human factors. The shoreline, where land meets the s...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Society of Land Measurements and Cadastre from Transylvania (SMTCT)
2025-03-01
|
| Series: | Nova Geodesia |
| Subjects: | |
| Online Access: | https://novageodesia.ro/index.php/ng/article/view/303 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Coastal areas are preferred home to about 40% of global population than any other ecosystems. Many organisms, including endangered species, live along sandy beaches. Nonetheless, such beaches are prone to erosion and flooding owing to natural and human factors. The shoreline, where land meets the sea, is highly dynamic and challenging to predict with accuracy. Existing predictive numerical models rely on multiple parameters, while statistical analysis of historical shoreline positions assume a linear distribution, overlooking the nonlinear nature of the data. This present study explored the application of nonlinear approaches like artificial neural networks (ANN) and other artificial intelligence techniques to predict shoreline change rate along the sandy beach of the study area using recurrent neural networks approaches: nonlinear autoregressive neural network (NARNN) and nonlinear autoregressive exogenous neural network (NARXNN); backpropagation neural network (BPNN), and compared results with multiple linear regression (MLR) model. Data used was partitioned into two: 70% was used for training and 30% was reserved for evaluating the performance of the models. From the developed models the output forecast for shoreline change were determined. NARXNN yielded best prediction, followed closely by NARNN and then BPNN as against MLR model. The optimum model developed for shoreline prediction provides invaluable information for planners and engineers of the coast.
|
|---|---|
| ISSN: | 2810-2754 |