WeSCoS Colloquium #10 “Study towards applying machine learning to numerical weather forecasting” By Takumi Honda (Associate Professor, Information Technology Center, The University of Tokyo)
WeSCoS Colloquium #10 “Study towards applying machine learning to numerical weather forecasting”
Presenter: Takumi Honda (Associate Professor, Information Technology Center, University of Tokyo)
Date and time: 3:00 pm-, August 6, 2024 (Tuesday)
Abstract:
Applications of machine learning are rapidly advancing in various fields, including numerical weather forecasting. In fact, some weather predictions made using machine learning models have achieved greater accuracy than traditional physical simulations. However, the approach of replacing all physical simulations included in numerical weather forecasts with machine learning models is currently not practical for on-site systems that require stable operation. Therefore, as a more conservative application of machine learning, we are proposing a new method that uses machine learning models only for quality control of observations [1]. This new method enables real-time estimation by using highly accurate predictions from machine learning as reference data when estimating the impact of observation assimilation on predictions. In this presentation, we would like to introduce a new method and discuss more proactive approaches, such as replacing the data assimilation process itself with machine learning.
[1] Honda, T., & Yamazaki, A. (2024). Machine learning enables real-time proactive quality control: A proof-of-concept study. Geophysical Research Letters, 51, e2023GL107938. https://doi.org/10.1029/2023GL107938
Venue: Zoom Pre-registration is required. Please register from here by August 5. We will send the Zoom URL on August 6.
·Language: Japanese
*The WeSCoS colloquium is part of the Moonshot-type R&D project Objective 8.
The colloquium in the core research ‘Control theory of weather-social coupled systems to support social decision-making’ is open to the public.
*WeSCoS stands for Weather-Society Coupling/Control System.