Improving wind-pressure relationship and intensity forecast skill of the operational HWRF

Hurricane Weather and Research Forecast (HWRF) model, which became an operational hurricane model from 2007 season, has shown comparable or better track forecast skill to other hurricane models (e.g. GFDL hurricane model). On the the other hand, the intensity prediction skill of HWRF is yet not as good as GFDL hurricane model. After analyzing the the results of HWRF, we found that most of intensity errors result from the weak intensity bias and wrong wind-pressure relationship of HWRF.

In order to improve the intensity forecast skills of HWRF, the tuning of horizontal diffusion coefficient and the modification of turbulent kinetic energy initialization (TKE) are performed. The experiments using reduced horizontal diffusion with zero initial TKE show promising results, which not only improve wind-pressure relationship and reduce the weak intensity bias of HWRF. These modifications are planing to implement to upgraded HWRF for the 2008 hurricane season.