The Hurricane Forecast Improvement Project

Abstract

Over the decade prior to 2007, the increasing vulnerability of the United States to damage and economic disruption from tropical storms and hurricanes was dramatically demonstrated by the impacts of a number of land-falling storms. In 2008, the National Oceanic and Atmospheric Administration (NOAA) established the Hurricane Forecast Improvement Project (HFIP) to significantly increase the agency's capability to address this vulnerability and begin to mitigate the impacts. In fiscal year 2009, The White House amended the president's budget and Congress appropriated funding to achieve a 20% reduction in forecast error (track and intensity) in 5 years with 50% reduction in 10 years. Over the past 3 years, HFIP has built computational infrastructure and implemented a focused set of cross-organizational research and development (R&D) activities to develop, demonstrate, and implement enhanced operational modeling capabilities to improve the numerical forecast guidance made available to the National Hurricane Center (NHC). HFIP collaborators, including federal laboratories and academic partners, have demonstrated potential for dramatic improvements in both hurricane track and intensity (up to 40%) prediction through the application of new techniques, including improved data assimilation, higher-resolution models (global and regional), enhanced model physics, better use of existing data sources to initialize regional hurricane models, and new postprocessing techniques. During each hurricane season, HFIP will run an experimental forecast system on NOAA's R&D high-performance computing to provide experimental improved guidance to NHC forecasters. Prior to each season, NHC will review and select a set of enhanced guidance products to evaluate operationally during the season (mid-July–October).