Over the past three decades, the average official NCEP/NHC 72 hour hurricane track forecast error has been reduced by over 50% which can be largely attributed to major advances in NWP guidance. Trend lines show that an acceleration in track forecast skill occurred in the mid-80's with increased use of the NCEP global model (the GFS, formerly the MRF/AVN) which made use of global data sources and continued with the operational implementation of the GFDL Hurricane Prediction System in 1995. The advancement of the GFS forecast skill again significantly increased during 2000 and 2001 compared to the previous five years on the order of 30% over the Atlantic basin and by as much 60% over the East Pacific basin in the GFS. Much of this improvement can be attributed to advanced techniques in data assimilation that has improved the overall tropical general circulation with increased utility of satellite data, the use of the NOAA G-IV dropsonde data and continuous model development including improved model physics. These GFS upgrades also benefitted the GFDL model track forecasts and in 200, a coupled GFDL model became operational at NCEP to begin to address improving intensity forecasts in hurricane scenarios where SST's are important to intensity changes.

Although though significant progress has been made in improving the overall track forecast skill, forecasting difficult track scenarios remain a challenge and the dynamical models do not show any reliable skill in intensity forecasts or hurricane related rainfall forecasts for landfalling storms which now account for most of the hurricane related fatalities.

It is anticipated that the Weather and Research Model for Hurricanes (HWRF) as another application of the community based WRF modeling system, will become operational at NCEP during 2006-2008 and will replace the GFDL model. The HWRF will be a movable nested high resolution coupled air/sea/land hurricane prediction system with advanced high resolution physics. The land surface coupling will take into account the changes in surface fluxes for landfalling storms and associated hydrology important to address the inland flooding problem. The HWRF will also make use of the operational NCEP wavewatch model to forecast hurricane generated wave heights and swells.

Although many challenges are inherent in developing this forecast system, one of the biggest scientific challenges resides in developing the next generation hurricane vortex initialization. In anticipation of meeting this challenge, NCEP is developing a local mesoscale data assimilation capability for initializing the hurricane core circulation making use of real time doppler radial velocities from the NOAA G-IV to describe the 3-D wind structure of the hurricane core circulation from the outflow layer to the surface. This capability is being developed in concert with NCEP's current 3-D var to make use of future advanced mesoscale variational data assimilation techniques being developed at NCEP. We believe this capability is critical to address the intensity problem by resolving the outflow layer of the hurricane core with detailed vertical resolution to the surface to properly simulate the boundary layer fluxes and resolve the vortex asymmetries.

NCEP anticipates a broad contribution from the overall tropical science/modeling community and looks forward to extensive collaboration in developing our next generation hurricane prediction system.

The GFDL Forecast System (Robert Tuleya)

A multiply nested regional forecast system was developed at NOAA’s Geophysical Fluid Dynamics Laboratory (GFDL) in the 1990’s. The forecast system was developed originally as a research tool to investigate the behavior and structure of tropical disturbances. It is the first model capable of simulating realistic wind intensity and meso-scale storm structure such as spiralbands. During its development, it has achieved a series of successful forecasts compared to other available models for Hurricanes Andrew(1992) and Emily(1193) and Tropical Storm Gordon(1994). The model became operational for the NWS beginning in 1995 and for the US Navy in 1996. Since that time it has been a premier tropical storm guidance product throughout the world.

The GFDL forecast system will be described including the model structure and the initialization method. In addition the post-processing model guidance will also be described. Both advantages and limitations of the GFDL forecast system will be discussed. Some time will be spent describing it’s evolution during computer changes and the relative role played by the GFDL model compared to other computer guidance.

  Back to Colloquia listing.