hurricanes

The DTC currently focuses its testing and evaluation of hurricane forecast systems on limited area mesoscale models with storm-following nesting capabilities. While the DTC does from time to time evaluate hurricane forecasts systems run by collaborating model groups, testing and evaluation by our Hurricane team generally focuses on configurations of the Hurricane Weather Research and Forecasting (HWRF) model run by the DTC. HWRF, a NOAA operational model for hurricane prediction, has several components (atmospheric model, coupler, ocean model, initialization and postprocessing), and the tests may emphasize one or more of these components. The alternate configurations for HWRF used in the tests can originate from NOAA/NCEP, from the DTC itself, or HWRF developers located in research laboratories and universities.

The ultimate goal of the DTC’s hurricane forecast systems testing and evaluation activities is to evaluate the performance of a new NWP technique for extended retrospective periods that encompass a variety of large scale atmospheric and oceanic regimes. Such extended period testing provides a robust data set for objective forecast verification, including the assessment of statistical significance. And yet, important information can also be gathered by considering case studies. Hence, some hurricane forecast system testing and evaluation activities may include a hierarchy of tests ranging from case studies to extended periods.