Weathernews' Abe: AI weather models gain ground on traditional forecasting

Weather forecasting has long depended on numerical weather prediction: supercomputers take observational data and simulate atmospheric conditions using the laws of physics. That approach demands enormous computing power and a deep physical model of the atmosphere.

In recent years, AI-based weather models have emerged as a different path. Instead of simulating physics, these systems analyze vast amounts of historical data to predict future weather patterns. According to the article, meteorological agencies and private-sector companies worldwide are now developing their own proprietary AI forecasting models. Weathernews executive officer Daisuke Abe explained one such model in Chiba in July.

The stakes come down to whether historical-data-driven models can match or complement the physics-based systems that agencies have relied on for decades. The outcome hinges on how well these AI models perform in practice across different weather situations, and on whether agencies and companies continue to invest in both approaches side by side.