The impact of a stochastic perturbation scheme on hurricane prognosis in NCEP global ensemble forecast

A Stochastic Perturbation Scheme was developed at NCEP to represent model related uncertainty and it will be implemented to the NCEP operational Global Ensemble Forecast System (GEFS). The added stochastic forcing in the model equations is based on total conventional tendencies of the ensemble perturbations. Extensive experiments showed that the scheme can significantly improve the general performance of the ensemble mean forecast and ensemble based probabilistic forecasts. Particularly, it increases the ensemble spread and improves ensemble distribution in the tropics.

Experiments for the period of August and September 2008 were performed using GEFS system with and without the stochastic perturbation scheme. In this study, we evaluate the impact of the stochastic perturbation scheme on the performance of hurricane prognosis in a global ensemble forecast setting, by comparing the two sets of ensemble forecasts. Emphasize will be on case studies of weather events with a hurricane moving towards and affecting North America, with track and intensity as the primary focus in the context of ensemble forecast.