sneha11naidu/Water_Quality_Prediction

This project is part of the EY AI & Data Challenge 2026, focused on building a robust machine learning model to predict water quality across various river locations in South Africa. The challenge involves predicting three critical water quality parameters using satellite imagery and climate data. # EY AI & Data Challenge 2026: Water Quality Prediction ## Project Overview This project is part of the **EY AI & Data Challenge 2026**, focused on building a robust machine learning model to predict water quality across various river locations in South Africa. The challenge involves predicting three critical water quality parameters using satellite imagery and climate data. ### Challenge Objectives - **Primary Goal**: Predict water quality parameters for river locations across South Africa - **Target Variables**: - **Total Alkalinity (TA)**: Measure of water's ability to neutralize acids - **Electrical Conductance (EC)**: Indicator of dissolved ions in water - **Dissolved Reactive Phosphorus (DRP)**: Key nutrient parameter affecting water quality - **Data Period**: 2011-2015 - **Study Region**: ~200 river monitoring locations across South Africa - **Secondary Goal**: Identify and interpret key factors influencing water quality variations --- ## Project Structure ``` ├── README.md # This file ├── requirements.txt # Python package dependencies │ ├── water_quality_training_dataset.csv # Training data with target variables ├── submission_template.csv # Template for predictions │ ├── Benchmark_Model_Notebook.ipynb # Main ML model and workflow │ ├── Landsat_Data_Extraction_Notebook.ipynb # Extracts Landsat satellite features ├── Landsat_Demonstration_Notebook.ipynb # Tutorial for Landsat data usage ├── landsat_features_training.csv # Pre-extracted Landsat features (training) ├── landsat_features_validation.csv # Pre-extracted Landsat features (validation) │ ├── TerraClimate_Data_Extraction_Notebook.ipynb # Extracts climate features ├── TerraClimate_Demonstration_Notebook.ipynb # Tutorial for TerraClimate data usage ├── terraclimate_features_training.csv # Pre-extracted climate features (training) └── terraclimat …