Enhancing effectiveness of numerical simulation with statistical techniques
【Notice】From April 1st, 2019, "Project" became "Center for Data Assimilation Research and Applications"
Numerical simulation is widely used in variety of fields such as weather prediction, fishery prediction, aeroplane designing, and building designing. Recent simulation models can compute the evolution of a system with high temporal and spatial resolution. However, numerical simulation just reproduces a scenario of temporal evolution of a system under given conditions. When conducting numerical simulation, it is crucial to give appropriate conditions such as initial conditions and boundary conditions. Statistical approaches are useful for appropriately setting the conditions to be given for simulation. In order to promote application researches of statistical techniques for enhancing effectiveness of numerical simulation, Support Project for Data Fusion Computation (SPDFC) has been launced.
Support Project for Data Fusion Computation (SPDFC) is a project aiming at providing our knowledge of novel statistical techniques for simulation researchers and conducting practical collaborative researches.
Available statistical techniques
- Data assimilation: estimating a scenario of temporal evolution by incorporating a sequence of the observational data into a numerical simulation model
- Statistical emulator: imitating a simulation model by a statistical model
Genta Ueno (Director) *1
Shunichi Nomura *1
Daisuke Murakami *1
Takashi Yamamoto *1
Institute of Mathematical Statistics *1
Shin'ya Nakano (Joint appointment)
National Institute of Polar Research *2
Yasunobu Ogawa (Joint appointment)
National Institute of Genetics *3
Akatsuki Kimura (Joint appointment)