PANCAST predicts thunderstorms across Africa using satellite data and artificial intelligence.
Covers the entire African continent with consistent monitoring.
Updated every 15 minutes, just a few minutes after receiving satellite data.
Forecasts up to 2 hours ahead at 30, 60, 90 and 120 minutes.
Provides detailed forecasts at around 3 km resolution.
PANCAST uses satellite observations to detect and track convective cloud systems in real time. These observations are processed to identify key features of developing storms, such as their location, size, and evolution over time.
A machine learning model analyses these features to estimate the probability of thunderstorm occurrence in the near future. The model is trained on historical data and designed to capture the behaviour of rapidly evolving convective systems.
The resulting forecasts are provided as probabilistic maps, allowing users to assess the likelihood of storms across large regions and at fine spatial resolution.