Over the space of only 3 years, artificial intelligence (AI) has revolutionised weather prediction, delivering models with comparable skill that run thousands of times faster than conventional models. The weather and climate community is currently retooling itself to understand and exploit this new science.
The computational speed-ups not only reduce the resources required to produce a weather forecast, but also allow for much larger ensembles to be run. This is a promising development for better quantifying the forecast uncertainty and capturing rare weather extremes.
While the current wave of AI models focuses on global forecasting, in AIGLE we plan to downscale global predictions to higher-resolution local predictions in Norway.
AIGLE will advance the application of AI weather and climate prediction at NORCE, focusing on model accuracy, alignment with physical principles, and their transformative potential for climate services.