Extreme weather forecasting technology

High-Accuracy Extreme Weather Forecasting Technology Tailored to Each Use Case through Efficient Observations and Improved Forecasting Models

Current research

In extreme weather forecasting, we develop accurate, real-time forecasting methods using observational data and numerical simulations, and identify the observational requirements needed to achieve this.

As global warming progresses, concerns are growing that damage from extreme weather will further increase. To reduce such damage, we need more accurate forecasts that utilize in situ observations over data-sparse ocean areas. This requires technologies and frameworks for efficiently collecting real-time in situ observational data at multiple locations and effectively incorporating them into forecasts.

We aim to address these challenges through the following initiatives. First, we will use numerical simulations to evaluate in advance the impact of observational data on forecasts, thereby improving the efficiency of the deployment and operation of actual observation instruments. Second, we will use the observational data obtained through our ultra-wide-area atmosphere and ocean observation technology (see related article), with appropriate quality checks, to improve forecast accuracy. Third, we will use in situ observations that simultaneously capture atmospheric and oceanic physical processes to clarify atmosphere-ocean interactions and incorporate the resulting knowledge into forecasting models. Finally, we will combine fast AI weather models with high-resolution physics-based models to achieve rapid, high-accuracy simulations.

Looking ahead, we will define forecasting requirements for each use case and deliver forecast information tailored to specific needs.