Data Assimilation and Predictability Studies for Improving Tropical Cyclone Intensity Forecasts
Lead PI: Dr. Takemasa Miyoshi, University of Maryland, College Park
Start Year: 2009 | Duration: 3 years
Partners: Naval Research Laboratory, Earth Simulator Center, National Central University, Taiwan, & Shanghai Typhoon Institute
In this project, an interdisciplinary team composed of experienced experts in the THORPEX Pacific Asian Regional Campaign (T-PARC) observations, in modeling, and in advanced data assimilation with ensemble Kalman filter (EnKF) will carry out experiments focused on the understanding and improving of the forecast of TC lifecycle evolution and intensity. This project focuses on both largescale environment and mesoscale phenomena in the TC system, which are major components responsible for intensity change, using several major codes already developed: the coupled ocean-atmosphere general circulation model (CGCM) known as CFES (CGCM for the Earth Simulator), the widely used mesoscale model known as WRF (Weather Research and Forecasting), and the LETKF (Local Ensemble Transform Kalman Filter) that is probably the most advanced data assimilation method for realistic systems and that was coded and tested by PI Miyoshi at Japanese Meteorological Agency (JMA). Because of high computational demands of this project, it is important to emphasize that all experiments with CFES will be carried out with the Earth Simulator, one of the largest supercomputer systems in Japan. Two major challenges in TC intensity forecasting are the general lack of observations in the vicinity of TCs and the adaptive representation of the forecast error covariance. In this project, both challenges will be addressed to improve TC intensity forecasting.