Skip to content

🌦️ WeDeAn – Weather Derivatives Analyzer

WeDeAn (Weather Derivatives Analyzer) is a Python and Streamlit-based application for analyzing weather data and valuing HDD/CDD-based weather derivatives. It integrates DWD (Deutscher Wetterdienst) datasets, computes degree-day indices, and applies burn analysis to estimate the theoretical payoffs and fair values of weather options.

Watch the video

👉 Check out on GitHub

🚀 Features

  • Load and process DWD weather datasets: Automatically reads and cleans station data (e.g., TMK, RSK, SDK) and prepares it for analysis.

  • Calculate Heating and Cooling Degree Days: Converts daily temperature data, the core indices used in weather derivatives.

  • Adjust for climatic trends: Optionally detrend long-term CDD/HDD series to remove warming or cooling biases.

  • Burn analysis pricing: Calculates historical payoffs and fair values of HDD/CDD-based options using actual observed weather.

  • Interactive visualizations: Displays time-series plots, rolling averages, payoff curves, and fair value results - fully integrated in a Streamlit dashboard.

📊 High-Level Process

  1. Load Dataset – Select a weather station and import its historical data (TMK, SDK, RSK, etc.).
  2. Clean & Filter – Restrict data to a selected observation window (Observation length) and remove years with incomplete data.
  3. Compute Degree-Day Indices – Aggregate daily HDDs/CDDs into seasonal accumulated values.
  4. Trend Adjustment (optional) – Remove long-term temperature trends.
  5. Calculate Payoffs – Compute annual option payoffs from accumulated indices.
  6. Price Options (Burn Analysis) – Estimate fair values as the discounted historical mean of those payoffs.

🧮 Core Concepts

HDDs (Heating Degree Days)

A measure of how much and for how long the outside air temperature is below a base temperature (commonly 18°C). Used to estimate heating demand - colder weather means higher HDDs and greater energy use.

Formula:

HDD_i = max(0, T_base – T_i)

CDDs (Cooling Degree Days)

A measure of how much and for how long the outside air temperature is above a base temperature (commonly 18°C). Used to estimate cooling demand - hotter weather means higher CDDs and more electricity use.

Formula:

CDD_i = max(0, T_i – T_base)

Fair Value of CDD/HDD Options

The expected (average) payout, discounted to present value.

Can be estimated by:

  • Historical simulation (burn analysis)
  • Statistical modeling
  • Monte Carlo simulation

Pricing Using Burn Analysis

A transparent pricing method using historical weather records.

  1. Compute the HDD/CDD index for each historical year.
  2. Calculate the option payoff (e.g., call/put) for each burn year.
  3. Take the average discounted payoff across all years as the fair value.

Assumption: Past temperature distributions represent the future (unless trend-adjusted).

▶️ Run the Application

1. Clone the repository

git clone https://github.com/trholy/wedean.git
cd wedean

2. Build and run docker container

docker-compose up --build

3. Open in your browser: The app will launch at http://localhost:8501

Example Output

  • Berlin Marzahn 2020
  • CDDs
  • Trend adjusted
  • Observation length: 12 years
  • Moving average length: 5 years
  • Interest rate: 2.0%
  • Tick size 100$

Description Description

📂 Project Structure

├── .dockerignore
├── .gitignore
├── Dockerfile
├── LICENSE
├── README.md
├── datasets
├── docker-compose.yml
├── pyproject.toml
├── setup.py
├── src
│   ├── __init__.py
│   ├── wedean
│   │   ├── calculation
│   │   │   ├── __init__.py
│   │   │   └── calculation.py
│   │   ├── data_handling
│   │   │   ├── __init__.py
│   │   │   └── data_handling.py
│   │   ├── plotting
│   │   │   ├── __init__.py
│   │   │   └── plotting.py
│   │   └── utils
│   │       ├── __init__.py
│   │       └── utils.py
└── streamlit-app
    ├── app.py
    ├── requirements.txt
    └── utils.py

📜 License

This project is released under the MIT License. You are free to use, modify, and distribute it with attribution.