Display Accessibility Tools

Accessibility Tools

Grayscale

Highlight Links

Change Contrast

Increase Text Size

Increase Letter Spacing

Readability Bar

Dyslexia Friendly Font

Increase Cursor Size

MICHIGAN STATE UNIVERSITY
College of Natural Science
IRES:ASSURE

Data Assimilation

DATA ASSIMILATION RESEARCH TEAM

LEAD PI / MENTOR: TAKEMASA MIYOSHI

This project aims at advancing data assimilation methods and their wide applications by integrating computer simulations and real-world data in the wisest way. Specifically, we tackle challenging problems to develop fundamental technologies integrating data-driven and process-driven approaches, taking advantage of advanced high-performance computers (HPC) and various sources of data in the new era of IoT (Internet of Things), Big Data and Artificial Intelligence (AI). Data Assimilation Research Team is strong in meteorological applications and numerical weather prediction, but the project is not limited to the specific application area.

The work of participating graduate students will advance the theory and application research on data assimilation, with potential new application areas depending on the students’ interests. Students will learn data assimilation and how to combine data-driven and process-driven approaches. Participants will also have a unique experience working on Fugaku, currently ranked #1 of Top500.

The general concept of combining process-driven simulations with data can lead to a bigger problem for a long-term collaboration. Specifically, participants working with U.S. based mentors experienced in HPC simulation, with research interest in data assimilation and its broader applications, theoretical developments for fusing data assimilation and AI will be a good fit for this project.

REQUIRED SKILLS

Technical skills for scientific computation such as shell scripting, FORTRAN/C++ coding, and data handling and visualization are required. For example, typical skills for numerical weather prediction include b-shell scripting, FORTRAN90 coding, and GrADS for data visualization, but these may be different in different application areas. AI and machine learning skills would be beneficial to explore new approaches combining data assimilation and machine learning. Experiences with HPC resources would be beneficial but not required.

COMPUTATIONAL RESOURCES

Fugaku Cluster servers