HAFS Ensemble Forecast for 2024 Atlantic Storms on Azure Cloud

The Hurricane Analysis and Forecast System (HAFS) Ensemble in Real-time on the AWS Cloud (HERC) experiment of 2023 was a highly successful real-time demonstration of the National Weather Service (NWS) hurricane model forecast. The verifications from the 2023 Atlantic hurricane season indicates that the HAFS ensemble system is under-dispersive for hurricane track and intensity forecasts. It also has the initial negative intensity bias due to the missing of the Vortex Initialization (VI) and Data Assimilations (DA). In 2024 hurricane season, the under dispersive issue is addressed by adding the new stochastic parameter perturbation, such as the ocean surface roughness perturbation. The other stochastic physics in HAFS ensemble includes Stochastically Perturbed Physics Tendencies (SPPT), Stochastically Kinetic Energy Backscatter (SKEB) and Stochastically Perturbed PBL Humidity (SHUM). The number of ensemble members is increased from 21 to 31 as well. The initial and boundary conditions are generated from the NCEP operational GEFS 31-member forecast dataset. Meanwhile, the operational HAFS-A initial vortex data with VI and DA is merged into the HAFS ensemble basin domain fields, which will significantly reduce the initial intensity error and bias. This new version of HAFS ensemble (version 2) including the VI/DA and stochastic parameter perturbations is expected to do a better job for the 2024 hurricane forecasts with the help of Azure Cloud Computing. The real time products from HAFS ensemble will provide hurricane probabilistic forecast guidance for NHC forecasters in 2024 hurricane season, including hurricane track, intensity, structure change, Rapid Intensification (RI) and hazards (wind, wave and storm surge) probability forecasts.