In this study, we evaluated the ability of the Numerical Tools for Hurricane Forecast (NTHF) system, operational at the Department of Meteorology of the Higher Institute of Technologies and Applied Sciences, University of Havana, Cuba, for forecasting the intensity and trajectory of the North Atlantic (NATL) tropical cyclones (TCs). To assess the ability of the NTHF system in the first five years (2016–2020) of operational runs, we used the best tracks from the National Hurricane Center HURDAT2 database. The errors in the track forecast increased linearly from 41 km at 6to 356 km at 120 h. In addition, NTHF underestimates the intensity of TCs, especially those that reach Category 3+ hurricanes on the Saffir–Simpson wind scale. Overall, the mean absolute error in forecasting the maximum wind speed (minimum pressure) ranged from 26.5 km/h (7 hPa) to 33.7 km/h (11.7 hPa). Moreover, the highest BIAS in the intensity forecast was found in the first 48 h, suggesting that model initialization is the main driver of NTHF errors in the forecast maximum wind speed and the minimum central pressure of TCs in the North Atlantic Basin.
Keywords: tropical cyclones; trajectory; intensity forecasts; NTHF; statistical validation; numerical model