YOSHIMURA, Hiroaki, Director
It goes without repeating that the global climate change accompanying warming of the planet is having negative impacts on natural ecosystems and is causing a variety of disasters in the societal sphere.The main cause is the emission of greenhouse gases, which is essentially a problem related to energy supply and consumption.The problems directly caused by climate change are related to natural phenomena, such as atmospheric circulation, ocean currents, and weather, as typified by heavy rainfall, typhoons, etc. The issues to be addressed by science are narrowed down to modeling of these phenomena, as well as prediction and control based on that modeling, and optimization of energy flows.Understanding weather, ocean currents, and atmospheric circulation phenomena requires consideration of the multiscale nature of physical phenomena, including both the familiar and the global scales, the chaotic complexity introduced by system nonlinearities, and the coupling with non-equilibrium thermodynamics involving mass transfer, in addition to continuum behavior.Furthermore, phase changes that occur on the microscopic scale appear as singular phenomena on the macroscopic scale, so it is necessary to view the microscopic and macroscopic scales as a large-scale structure that is connected by time and space.Such problems are underpinned by classical physics, especially classical field theory and the theory of continuum mechanics, such as fluid mechanics and thermodynamics, and the analytical mechanics and statistical mechanics that describe them. They have been treated in mathematics as nonlinear infinite-dimensional dynamical systems on manifolds such as Lie groups.However, the conventional mathematical approach has been to analyze physical phenomena by breaking them down into individual elements, simplifying them, and constructing mathematical models of them.Yet, it is difficult to accurately capture complex phenomena that exist in the real world: multiphysics modeling becomes essential.There is also the additional requirement to realize modeling as a large-scale system using vast amounts of data from actual phenomena and experiments, along with multi-scale mathematics linking micro and macro scales, and to aim for prediction, control, and optimization of energy flow as the final outputs.The basic techniques are numerical analysis and engineering optimal design, but discrete mathematics such as stochastic differential equations, structure preserving computation, and network theory are expanding as mathematical frontiers, and integration with the latest developments in computational science, such as the data sciences of AI, machine learning, and information geometry, is essential.Our research aims to create new values and seeks to solve these problems through systematic research by specialists in various fields, from the mathematical science standpoint.