Tackling Climate Change with Time Series Analysis and Forecasting

This repository provides the code, experiments, and data samples for a series of articles exploring how time series analysis can address critical climate change challenges.

The project is divided into 8 distinct tasks, ranging from renewable energy forecasting to extreme event detection.

Project Content & Articles

Each module includes a dedicated implementation folder and an accompanying article explaining the methodology and impact.

| Task | Problem Domain | Forecasting Topic | Article Link | | :--- | :--- | :--- | :--- | | Part 1 | Wind Power | Univariate Forecasting | Read Article | | Part 2 | Solar Radiation | Multivariate Forecasting | Read Article | | Part 3 | Ocean Wave Height | Exceedance Probability | Read Article | | Part 4 | Energy Load | Seasonality Management | Read Article | | Part 5 | Extreme Weather | Event Detection | Read Article | | Part 6 | Dew Point | Deep Learning | Read Article | | Part 7 | Food Demand | Time Series Clustering | Read Article | | Part 8 | Origin-Demand | GPS & Demand Forecasting | Read Article |

📊 Datasets

All datasets used in these tasks are sourced from public repositories.

• Full Data: Please refer to the specific Medium article for each part to find the original data source and licensing information.
• Samples: This repository contains small data samples in each subfolder to ensure the code is runnable and to demonstrate the required data formats.

Getting Started

1. Navigate to the folder of the task you are interested in (e.g., content/part_1).
2. Read the corresponding article to understand the context and theoretical approach.
3. Run the scripts provided within the folder to reproduce the analysis.