An Examination of Model and Official National Hurricane Center Tropical Cyclone Size Forecasts

Abstract

Abstract While the National Hurricane Center (NHC) has been issuing analyses and forecasts of tropical cyclone wind radii for several years, little documentation has been provided about the errors in these forecasts. A key hurdle in providing routine verification of these forecasts is that the uncertainty in the wind radii best tracks is quite large for tropical cyclones that are well away from land and unmonitored by aircraft reconnaissance. This study evaluates the errors of a subset of NHC and model 34-, 50-, and 64-kt (1 kt = 0.514 m s−1) wind radii forecasts from 2008 through 2012 that had aircraft reconnaissance available at both the initial and verification times. The results show that the NHC wind radii average errors increased with forecast time but were skillful when compared against climatology and persistence. The dynamical models, however, were not skillful and had errors that were much larger than the NHC forecasts, with substantial negative (too small) biases even after accounting for their initial size differences versus the tropical cyclone’s current wind radii. Improvements in wind radii forecasting will come about through a combination of better methods for observing tropical cyclone size as well as enhanced prediction techniques (dynamical models, statistical methods, and consensus approaches).