Tech giants like Google, Nvidia, and Huawei are revolutionising weather forecasting with their Artificial Intelligence (AI) models, a study revealed on Monday.
Published in the prestigious journal npj Climate and Atmospheric Science, the study demonstrates how AI-powered weather prediction models can swiftly and precisely anticipate the trajectory and intensity of major storms. These AI-based forecasts, according to researchers, match the accuracy of traditional methods while being faster, cost-effective, and demanding fewer computational resources.
The research, led by Professor Andrew Charlton-Perez, focuses on the November 2023 Storm Ciaran that wreaked havoc across northern and central Europe. The team analysed the storm's behaviour using advanced AI models developed by Google, Nvidia, and Huawei, comparing their performance with traditional physics-based models.
Remarkably, the AI models accurately forecasted the storm's rapid intensification and trajectory up to 48 hours in advance. According to the researchers, the forecasts were nearly identical to those generated by conventional methods.
Professor Charlton-Perez commented, "AI is transforming weather forecasting before our eyes. Two years ago, modern ML techniques were rarely being applied to make weather forecasts. Now, we have multiple models that can produce 10-day global forecasts in minutes."
The study highlights the AI models' ability to capture crucial atmospheric conditions that drive a storm's development, including its interaction with the jet stream—a narrow corridor of strong high-level winds.