［ 2017年02月27日 ］
RIKEN International Symposium on Data Assimilation 2017
"Mesoscale hybrid data assimilation system based on JMA nonhydrostatic model"
We develop an adjoint-based four-dimensional variational (4D-Var) data assimilation system that employed a background error covariance matrix B constructed from the NMC method and perturbations in a local ensemble transform Kalman filter (LETKF) system. The system is based on the Japan Meteorological Agency's nonhydrostatic model. The assimilation of a pseudo-single observation of sea level pressure located at a tropical cyclone (TC) center yielded analysis increments physically consistent with what is expected of a mature TC in the hybrid system at the beginning of the assimilation window, whereas analogous experiments performed using the conventional 4D-Var system did not. At the end, the structures of the 4D-Var-based increments became similar to one another, while the analysis increment by the conventional 4D-Var system was broad in the horizontal direction. Realistic experiments showed that the hybrid system provided initial conditions yielding more accurate TC track and intensity forecasts than those achievable by the conventional 4D-Var system. The hybrid system also yielded some statistically significant improvements in forecasting local heavy rainfall events in terms of fraction skill scores when a 160 km x 160 km window size was used.
|名前:Kosuke Ito||所属:University of the Ryukyus|