Google is rolling out AI-powered weather forecasts across Africa, offering a new level of precision in short-term precipitation predictions for a continent where radar coverage is sparse. The initiative, powered by Google Research’s AI-driven nowcasting model, MetNet, aims to provide real-time weather updates that could benefit millions, from urban commuters to farmers and coastal fishing communities.
“The forecasts are made possible through advancements in Google Research’s AI-powered nowcasting model MetNet, which uses satellite data and ground observations to produce state-of-the-art precipitation forecasts in data-sparse regions of the world,” said Emmanuel Asiedu Brempong and Shreya Agrawal, research engineers at Google Research. “Google’s technology can predict global precipitation with high accuracy within a 5km radius every 15 minutes for the next 12 hours — all under one minute.”
The technology, now live on Google Search across the continent, delivers short-term rainfall predictions within a 5-kilometer radius, updating every 15 minutes for up to 12 hours. Google says MetNet can generate these forecasts in under a minute, using global satellite observations to compensate for Africa’s limited weather radar network.
Bridging Africa’s weather data divide
Accurate weather forecasting remains a challenge across much of the continent. While North America has 291 radar stations tracking precipitation, Africa has just 37, making it difficult to provide reliable short-term forecasts. Traditional global weather models, which rely heavily on ground-based data, often struggle to generate precise predictions for Africa’s diverse climates.
Google’s approach leverages satellite imagery and machine learning to analyze atmospheric patterns and predict rainfall with greater accuracy. The system is the first of its kind to incorporate real-time observations from multiple satellites, allowing it to generate forecasts even in areas with no local radar coverage.
This advancement could be transformative for farmers, who rely on precise weather predictions to plan planting and harvesting schedules, as well as for coastal communities where sudden storms pose risks to fishing activities. It could also help city dwellers better prepare for daily weather changes, reducing disruptions caused by unpredictable rain patterns.
AI-powered forecasts now available across Africa
The AI-powered nowcasting feature is now integrated into Google Search, meaning users across Africa will see highly localized, real-time precipitation forecasts when searching for weather updates. A demonstration in Kenya showed the model adjusting forecasts dynamically based on incoming satellite data, providing updates with a level of granularity not previously available in the region.
Google is also working with scientific communities across Africa to refine the model and improve forecast accuracy. “We are excited to see this work leave our offices in Accra and Nairobi and make its way to users all across Africa. We are engaging with local scientific communities to further evaluate and improve forecast accuracy globally. We’re also looking to launch the technology in additional regions with limited radar.”
The rollout is part of Google Research Africa’s broader efforts to apply AI to real-world challenges on the continent. Other initiatives include Open Buildings, a project mapping structures to improve disaster response and urban planning, and various machine learning applications aimed at climate resilience.
As extreme weather events become more frequent due to climate change, experts say better forecasting tools could be crucial in helping African communities prepare. Google’s AI-driven approach represents a step toward a future where real-time, high-precision weather predictions are accessible to millions—whether they’re checking if they need an umbrella or planning their next fishing trip.
