Machine learning-based assessment of flood susceptibility in the Eastern Mediterranean: a case study of Baniyas River basin
Floods are one of the main damaging physical catastrophes, inducing economic losses and human casualties worldwide. The Eastern Mediterranean region is exposed to devastating flood events annually, with catastrophic consequences due to the complexity of the physical and human geographical characteri...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2025-12-01
|
| Series: | Geomatics, Natural Hazards & Risk |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/19475705.2025.2524417 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Floods are one of the main damaging physical catastrophes, inducing economic losses and human casualties worldwide. The Eastern Mediterranean region is exposed to devastating flood events annually, with catastrophic consequences due to the complexity of the physical and human geographical characteristics. In this analysis, the performance of four ensemble machine learning algorithms (ML), that is support vector machine (SVM), random forest (RF), artificial neural network (ANN) and extreme gradient boost (XGBoost), was compared and tested in mapping flood susceptibility in the Eastern Mediterranean region. In the Baniyas River basin in western Syria, 1,100 flood events with 20 flood-driving factors were relied upon to achieve the goal of this assessment. The multicollinearity test results showed that all selected factors can be incorporated into the modelling process. Additionally, all the applied algorithms showed reliable and accurate performance in modeling flood susceptibility; however, the XGBoost algorithm achieved the strongest performance compared to other models with an AUC value of 0.98. Overall, the current results guide urban planners and land managers in increasing the quality of sustainable development practices and enhancing community resilience to flood risk in the study area. |
|---|---|
| ISSN: | 1947-5705 1947-5713 |