Using Novel Observations at the Air-Sea Interface to Evaluate the Coupled Ocean-Atmosphere Background Ensemble Forecasts from a Self-Cycled Hafs Data Assimilation System for Hurricane Finoa (2022)

A field campaign was conducted during Hurricane Fiona (2022) where unique observations sampling the properties at the air-sea interface were collected. These observations included saildrone, and paired dropsonde and AXBT. Saildrone measured air-sea interface variables including atmospheric parameters and oceanic parameters (e.g., wind, air temperature, water temperature, and salinity). Dropsondes provided vertical profiles in atmospheric boundary layer. Paired dropsonde and AXBT allowed measures of boundary layer structure and ocean thermal structure data to depths of approximately 400 meters. These observations altogether could also derive 3D thermal fluxes at the air-sea interface. A self-cycled data assimilation experiments were designed and implemented for Hurricane Analysis and Forecast system (HAFS) coupled with HYbrid Coordinate Ocean Model (HYCOM) from the tropical storm phase of Fiona (2022). These novel observations at the air-sea interface were leveraged to evaluate the background ensemble of this DA system. It was found that the deterministic forecast matched the dropsonde/AXBT observations well in both boundary layer and ocean structure, performing even better than the 2022 HAFS-A operational run. Air-sea interface verifications from saildrones indicated good agreement between the background forecast and observations, with some identified biases. Thermal flux verifications also demonstrated consistency in terms of the storm shape and size. Results from the background ensemble verification against the saildrone and the dropsonde boundary layer observations showed that the ensemble spread was overall under-dispersive. This study shed lights on the performance of the coupled ocean-atmosphere background ensemble and provide insights on how to improve the current state of art DA system for the eventual fully coupled ocean-atmosphere DA for hurricane prediction.