Global Environmental Science
Prediction of Global Change
In recent years, reliable predictions are provided for global warming caused by the increasing anthropogenic greenhouse gases.
More precise regional climate prediction is demanded, however, for mitigation on human impact especially for agriculture in each country or region.
Numerous technical problems then arise, and computation becomes enormous when the model resolution is increased for regional climate prediction.
In our group, the prediction of global change is performed by a combination of the regional climate model nested in the global climate model.
The figures illustrate the precipitation change (blue for decrease) between 1990s and 2070s over the world predicted by the general circulation model at the Meteorological Research Institute (lower) and over the east Mediterranean where severe drought is anticipated, predicted by the regional climate model at the University of Tsukuba (upper).
|
Data Assimilation for Meteorological Observations
Model atmosphere would approach the real atmosphere and eventually become identical if observed data are assimilated into the model atmosphere for infinite time interval. The technique is called 4 dimensional data assimilation. By this technique, we can infer the global feature of the atmosphere even if the meteorological observations are available sparsely only on the land. In our group, comparative experiments are conducted for the products of 4 dimensional data assimilation by many operational meteorological centers over the world. A new assimilation technique is developed by the application of Kalman Filter in the ensemble forecasts.
Biological Science
Phylogenetic Inference on the Universal Tree of Life
The central focus of our research is to gain insight into origin and early evolution of eukaryotes, the most important open problem in evolutionary biology.One of the goal of this research is to reconstruct the Universal Tree of Life including diverse organisms on the earth, based on molecular phylogenetic approach. Since a reliable phylogenetic inference simultaneously requires numerous genes and species, a large scale data analysis using high performance computers plays a key role in our study.
|
