Hurricane Initialization in HWRF Model

Qingfu Liu
2006-01-01
Abstract:The GFDL Hurricane Prediction System has been the operational hurricane forecast model at the National Centers for Environmental Prediction (NCEP) since 1995. The initialization of hurricanes in the GFDL model uses a vortex replacement strategy, which consists of three major steps: 1) interpolate the global analysis fields from Global Forecast System (GFS) onto the operational GFDL hurricane model grid; 2) remove the GFS vortex from the global analysis; and 3) add a high resolution, model-consistent vortex (Kurihara, et al. 1995). The GFDL model has been a star performer in terms of track forecast since its operational implementation. Even as a high resolution model, it still lacks forecast skill in intensity. As more observational data become available in the hurricane region, it is crucial to use those data in the hurricane initialization to improve the track and intensity forecasts. However, the GFDL model does not have its own data assimilation package and can’ t use those data to generate a more realistic vortex structure in its hurricane initialization. In order to utilize the observational data in the hurricane analysis, the Hurricane WRF (HWRF) project was created in 2000. The goal of this project is to replace the GFDL model as the operational hurricane model in 2007. The HWRF project takes advantage of WRF (Weather Research and Forecast) model development, and adds unique features (such as the moving nest and hurricane data assimilation) for hurricane prediction. Significant progress in the HWRF project has been made in the areas of model development and data assimilation during the last four years. This paper reviews only the progress in the HWRF data analysis.
What problem does this paper attempt to address?