AI TECHNOLOGY
Imagine getting accurate weather forecasts without the need for massive supercomputers or teams of meteorologists.
That’s exactly what Aardvark Weather, a new AI-powered model, is set to do, delivering forecasts faster, cheaper, and with far less computing power than traditional methods.
AI reads the skies
Instead of crunching numbers for hours on specialised systems, Aardvark learns from raw data collected by weather stations, satellites, planes, and ships to predict conditions in real-time.
This means forecasts could be quicker, more precise, and much more cost-effective.
AI vs. traditional forecasting:
AI models like Aardvark use just 10% of the data that conventional systems require but can still outperform some US forecasting tools.
The approach builds on AI research by Google, Microsoft, and Huawei, proving that AI can replace parts of traditional weather modelling.
If widely adopted, this could democratise forecasting, making accurate predictions available to more countries, industries, and policymakers.
AI could extend forecast accuracy from five to eight days, while also making hyper-local predictions for industries like farming and renewable energy.
It could even help emergency services plan ahead for hurricanes, wildfires, and extreme weather events.
If it’s trained on unpredictable, unhinged UK weather, it might become the first AI to experience stress.