Research & Methodology

How we use Markov Chains and Satellite Imagery to predict weather stability.

The Markov Chain Model

Our "AI Stability Engine" utilizes a probabilistic Markov Chain model to predict short-term weather changes. By analyzing the current weather state (e.g., Clear, Cloudy, Rain), the model calculates the probability of transitioning to a different state within the next 3 hours.

Key Metric: Risk Score. A score above 50% indicates a high instability and potential for rain.

Data Sources

  • Open-Meteo API: Provides real-time and historical weather data.
  • OpenStreetMap (Nominatim): Powers our geocoding and reverse-geocoding services.
  • Google News RSS: Delivers real-time climate news updates.