Thursday, 3 April 2014: 2:00 PM
Regency Ballroom (Town and Country Resort )
Tropical Cyclone (TC) forecasting is a multi-scale problem. In order to predict TC structure, size and intensity changes with fidelity, the inner core structure of these storms should be resolved at 1-3 km horizontal grid spacing. In addition, TC systems are also moving entities that can travel up to thousands of kilometers within a forecast window. During the course of movement, TC systems can interact with terrains, fronts that create shear, the ocean, and even other neighboring TC systems within a certain range. Such multi-scale interactions could further amplify the forecast uncertainty of TC track, intensity, and structure if not properly represented in the modeling system. With the support of NOAA's Hurricane Forecast Improvement Project (HFIP), we have developed a basin-wide HWRF modeling system with multiple movable nests based on the NCEP's operational HWRF modeling system. The modeling system has the potentials to better represent TC multi-scale interactions: i.e., large-scale environment, across scale initialization, storm-storm interactions, land-ocean contrast effect, etc. To demonstrate the effectiveness of the system, we first verified the forecast track and intensity errors by this system in the Atlantic and East Pacific basins on 2012 retrospective forecasts and 2013 real-time forecasts. The results show both intensities and tracks are either superior or comparable to those from current operational HWRF in most of forecast leading times. Then we further illustrate sensitivities of terrain interaction and initialization at different scales represented by this system. The multi-scale interactions particularly of the storm-storm interactions are also examined in this study.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner