제목: AI/ML in Weather Prediction​​​

​​

​​​​​​​​​​​​​​​​​​연사: 오태진 박사 (KIAPS)​

​

장소: 과학관 B102​호​​

Abstract:

​​Artificial intelligence and machine learning (AI/ML) are emerging as powerful tools in numerical weather prediction (NWP). This talk will survey the application of deep learning to weather forecasting, covering foundational concepts, current applications, and future

directions.

We will begin with a brief overview of the core deep learning architectures, such as Convolutional Neural Networks and Transformers, that have enabled progress in processing complex, spatio-temporal atmospheric data. The presentation will then focus on the current generation of data-driven weather models, which learn atmospheric dynamics directly from reanalysis data. We will discuss how these models achieve competitive forecast skill while significantly reducing the computational cost compared to traditional NWP systems.

Finally, the talk will address key areas of future work. This includes challenges and opportunities in high-performance computing, the creation of integrated AI-based model systems, and the potential of hybrid approaches that merge physics-based simulations with

machine learning.