pchukwuemeka424/Machine-Learning-and-Morphometric-Analysis-for-Runoff-Dynamics-Enhancing-Flood

Flooding is a recurring environmental hazard with devastating socio-economic and ecological impacts, especially in vulnerable regions like Bayelsa State, Nigeria. The state’s low-lying terrain, dense river networks, and poor drainage infrastructure exacerbate its flood susceptibility. --- # Flood Susceptibility Mapping Using Machine Learning and Morphometric Analysis ## Overview This project integrates **Machine Learning (ML)** techniques and **morphometric analysis** to assess flood susceptibility across four major catchments in Bayelsa State, Nigeria. The study leverages **Shuttle Radar Topographic Mission (SRTM)** data to extract hydrological features and applies machine learning classifiers to classify flood-prone zones. The goal is to enhance flood prediction accuracy and granularity beyond traditional morphometric methods. ## Dataset The dataset used for this analysis includes key morphometric features, such as: - Drainage Density (Dd) - Stream Frequency (Fs) - Relief Ratio (Rh) - Bifurcation Ratio (Rbm) - Infiltration Number (If) - Circularity Ratio (Ff) - Hypsometric Integral (Hh) - Average Slope (Slope) The flood-prone labels are generated from historical flood records, with each catchment marked as either **flood-prone (1)** or **not flood-prone (0)**. You can access the dataset template for training in `morphometric_ml_dataset.csv`. ## Requirements Ensure you have the following Python packages installed: - `pandas` - `numpy` - `scikit-learn` - `matplotlib` - `seaborn` - `xgboost` - `lightgbm` - `shap` You can install them via `pip`: ```bash pip install pandas numpy scikit-learn matplotlib seaborn xgboost lightgbm shap ``` ## File Structure ``` ├── flood_susceptibility_ml.py # Enhanced Python script for model training and evaluation ├── morphometric_ml_dataset.csv # Dataset template for morphometric features ├── results/ # Directory for generated results and visualizations │ ├── confusion_matrices.png # Confusion matrices for all classifiers │ ├── roc_curves.png # ROC Curve comparison figure │ ├── model_comparison.csv # Performance comparison table │ ├── cross_validation_results.csv# 5-fold cross-validation results │ ├── model_predictions.csv # Predictio …