Second IMT-Atlantique & Kyoto University & RIKEN joint Data Assimilation workshop
Overview
- Date: 10:30-17:00 29th October 2024
- Place: Kyoto University, Department of mathematics, Building #3 Room 3-152 (See access information here)
- Language: English
Registration
- Registrations are now closed.
- The deadline for registration is October the 15th 2024.
Overview
Data Assimilation (DA) is closely related to Uncertainty Quantification (UQ) in applied mathematics and is an interdisciplinary science that combines mathematical models (mathematical science), simulations (computational science), and observations (data science). Its theoretical background includes dynamical systems, probability, statistical mathematics, differential equation, numerical analysis, and data science, and it has been applied to not only meteorology but also various research fields. Moreover, Machine Learning (ML) has been making rapid progress, and various achievements in geoscience fields have been made such as application to precipitation forecasting and model bias correction.IMT-Atlantique has its strength in statistical mathematics, ML, and applications to oceanography, and signed an MOU with RIKEN in 2019 to enhance the theoretical aspects of DA. IMT-Atlantique and RIKEN have been actively working together to organize joint workshops, organize sessions at an international conference, and send graduate students from IMT-Atlantique to RIKEN. In addition, Kyoto University and RIKEN have regularly organized joint data assimilation workshops focusing on mathematical science and theory since 2013. To extend these collaborations, we organize the IMT-Atlantique & Kyoto University & RIKEN joint data assimilation workshop at this time.
Tentative Program (To be updated)
| Time | Speaker | Title |
|---|---|---|
| 10:00-10:30 | Registration | |
| 10:30-10:35 [Chair: S. Otsuka] |
T. Miyoshi, P. Tandeo, and T. Sakajo |
Opening remarks |
| 10:35-11:35 | Alberto Carrassi (University of Bologna) |
Machine learning and its merge with data assimilation to discover sub manifold and instabilities in nonlinear systems |
| 11:35-12:35 | Lunch Break | |
| 12:35-13:30 | Juan Ruiz (CIMA) |
Machine learning and iterative approaches for non-linear filtering and smoothing |
| 13:30-14:00 | Takeshi Enomoto (Kyoto University) |
Jacobians and adjoints using automatic differentiation |
| 14:00-14:30 | Saori Nakashita (Kyoto University) |
Flow-dependent large-scale blending for limited-area ensemble data assimilation |
| 14:30-14:45 | Coffee Break | |
| 14:45-15:15 | Juan Martín Guerrieri (National University of Northeast, Corrientes) |
Local Mapping Particle Filter |
| 15:15-15:45 | Michael Goodliff (RIKEN) |
Using Data Assimilation to Improve Data Driven Surrogate Models |
| 15:45-16:15 | Erwan Oulhen (IMT Atlantic) |
Ocean state reconstruction using the analog method and in situ observations |
| 16:15-16:45 | Oscar Chapron (IMT Atlantic) |
Optimal Sensor Placement Strategies Using Gumbel-Softmax for Deep-Learning Ocean Reconstruction |
| 16:45-16:55 | Discussions | |
| 16:55-17:00 | T. Miyoshi, P. Tandeo, and T. Sakajo | Closing remarks |
Past events
Organizer
Inquiry
- RIKEN Center for Computational Science (R-CCS) Data Assimilation Research Team
- Email: da-joint-staff [at] ml.riken.jp (*Replace [at] with @)
Protection of personal data
In accordance with RIKEN's rule on protection of personal data, any data that we gain through our website will be protected strictly and will never be released, given, or lent to any third party without just cause.