NOAA established the Hurricane Forecast Improvement Project (HFIP) to develop a 10-year strategy to improve hurricane forecasts. Tropical cyclone probabilistic prediction is very important to the decision-making, such as the evacuation orders for reducing the risk to life and property in the coastal regions. The efforts will be achieved by the ensemble based numerical prediction systems.

This study is focused on tropical cyclone track forecast based on the multiple-model ensemble forecast system. We use five global ensemble forecast systems, which are being run in National Center of Environmental Prediction (NCEP), Fleet Numerical Meteorology and Oceanography Center (FNMOC), Canadian Meteorology Center (CMC), European Center for Medium range Weather Forecasting (ECMWF) and UK's National Weather Service (Met Office). The global tropical cyclone track probabilistic forecast is done by the combination of the TC-ensemble forecasts in the five centers. The results show encouraging improvements for hurricane /typhoon track prediction through this multiple global-model ensemble forecast system.