🌍 Climate Change Modeling with Machine Learning

This repository contains an end-to-end data science project focused on modeling and analyzing climate change indicators using real-world climate data and user sentiment. The project combines Natural Language Processing (NLP), exploratory data analysis, machine learning, and geospatial visualizations to assess and project climate change impacts.

📌 Objectives

• Predict and analyze climate change indicators such as temperature anomalies, CO₂ emissions, and extreme weather trends.
• Analyze public sentiment around climate change using Facebook comments from NASA's climate page.
• Apply machine learning to forecast climate metrics and visualize their progression over time and geography.

📊 Tech Stack

• Language: Python
• Libraries: pandas, numpy, matplotlib, seaborn, scikit-learn, folium, geopandas, tqdm
• ML Models: Random Forest, XGBoost, LSTM (future scope)
• Development Tools: Jupyter Notebook, VS Code

🧠 Features

1. Exploratory Data Analysis (EDA):

• CO₂ emission patterns across time and location
• Sentiment trends on climate change topics
• Outlier detection, skewness handling, and missing value imputation

3. Sentiment Analysis (NLP)

• Analyze over 500 user comments from NASA’s Facebook Climate Change page (2020–2023).
• Perform trend analysis and topic modeling using NLP.
• Data privacy is preserved using SHA-256 hashing for user anonymity.

4. Climate Data Modeling

• Dataset includes climate metrics such as CO₂ levels, solar radiation, temperature, sea level, and more.
• Feature engineering and preprocessing (normalization, encoding, handling outliers and missing values).
• Advanced time-series visualizations by week, year, and geographical coordinates.

5. Machine Learning

• Trained multiple models: Random Forest, Gradient Boosting, Neural Networks, and LSTM.
• Model evaluation using MAE, MSE, R², and cross-validation.
• Future forecasting and scenario simulation.

🗂️ Project Structure

├── main.ipynb              # Main notebook for climate modeling
├── data/                   # Raw and cleaned datasets
└── README.md               # Project documentation

Set Up Environment:

• Recommended: Create a virtual environment

    python -m venv venv
    source venv/bin/activate  # or venv\Scripts\activate
    

• Install required packages:

    pip install -r requirements.txt

• Run the Notebook:

    Open main.ipynb in Jupyter Notebook and run the cells step-by-step.

📌 Key Insights

• Public sentiment shows increasing concern and awareness about climate change.

• High emissions correlate with specific aerosol and cloud indicators.

• Significant seasonal and geographical emission trends were discovered.

📈 Future Improvements

• Integrate real-time data updates using APIs.

• Deploy model as a Flask/Django web app.

• Collaborate with domain experts for deeper scientific validation.

🙏 Acknowledgements

• NASA Climate Change Facebook for social sentiment data

• NOAA, IPCC, and Kaggle for climate datasets

• BriantOliveira for reference EDA structure

📬 Contact

Amrutha C

GitHub: @amruthadevops