Hurricane ensemble prediction using NCEP operational HWRF system and GEFS perturbations

This study illustrates the capability of the ensemble technique to improve hurricane track and intensity forecasts using the operational Hurricane Weather Research and Forecasting (HWRF) system at EMC.

In order to study the sensitivity of HWRF model forecasts to perturbations of the initial large-scale flows and lateral boundary conditions, the forecast data from NCEP Global Ensemble Forecast System (GEFS) are used in the experiment. The GEFS data includes 21 ensemble members from T126L28 model runs. A full life cycle of hurricane Hanna (2008) is conducted for 5-day forecasts.

The results show that the hurricane ensemble forecast tracks are very sensitive to the initial large scale flows, and spread around the observed track. The ensemble prediction method leads to an improvement in the hurricane track forecasts. The track position errors are largely reduced by the ensemble mean in total of 40 cases. The ensemble intensity forecasts remain same forecast skills as single deterministic forecast, possibly because the single control run uses higher resolution (T382L64) GFS data as its initial condition. It is also noted that the spread of the ensemble track forecasts is useful to assess the reliability of the predictions.