Abstract
Tropical cyclone (TC) genesis is a process by which some synoptic-scale or mesoscale weather feature in the tropical areas take on the characteristics of a tropical cyclone. Many TC genesis
forecast products produced by numerical models have been developed to provide guidance for forecasters.
This project has studied the difference including the genesis time and locations between the actual tracks of North Atlantic tropical cyclones and the predictions of Hurricane Analysis and Forecast System (HAFS) model during the 2019-2021 hurricane seasons. The contingency table was used to compare with actual genesis events in the study period. For the cases that a tropical cyclone formed within 24 hours by the forecast model, the numerical model genesis points were compared with the actual genesis locations. The main error range of time or distance between the HAFS numerical model and the actual track were analyzed. At the same time, individual cases that have significant impact on Florida were selected for further analysis.
The results of above analysis showed that the error of HAFS model for the hurricane genesis location were mainly within 200 km and the error of the genesis time were mainly within 36 hours. With the approaching of initial prediction time, the error showed a trend of gradually narrowing. In addition, the results may provide valuable information for forecasters on predicting TC genesis in the future hurricane seasons as well as early warning to the public.