Naonori Ueda
Naonori Ueda works on machine learning for disaster resilience. As of October 2026, Ueda directs RIKEN’s Disaster Resilience Science Team. He coauthored a 2023 study using a generative adversarial network to emulate atmospheric forces for storm-surge modeling. Its evaluation used historical typhoons. His 2024 work includes deep inpainting and neural surrogate models for seismic ground motion. Earthquake prediction and improved disaster recovery remain research goals.
Timeline
-
2023
Modeling typhoon forces
Ueda coauthored a study using generative adversarial networks to improve the wind and pressure inputs used in storm-surge models. The method translates a fast parametric model’s output toward the structure of numerical weather predictions. This links learned models with physical simulation.
-
2023
Testing historical storms
The study used 34 historical typhoon events for training and evaluated storm-surge simulations for four later events. The authors reported improved accuracy over the parametric model in that setting.
-
2024
Research on seismic motion
Ueda coauthored deep inpainting research for spatial interpolation of seismic motions. He also coauthored a neural-network surrogate model for seismic ground motion. These studies address how to model spatial hazard information with learning methods.
-
As of October 2026
Disaster research at RIKEN
Ueda directs RIKEN’s Disaster Resilience Science Team. The team aims to limit damage and support recovery after disasters. Its research subjects include autonomous learning systems, earthquake models and ground-motion prediction.
Sources and credits
Photo: www.riken.jp; creator not stated. License: Not stated by source. Source: riken.jp. Cropped to a square and resized. Ask us to remove or replace this photo.