May 22 - July 30
Evaluating the Vertical Mixing Schemes in the Ocean Surface Boundary Layer in the Intensification Forecast of Hurricane Fiona (2022)
Alexis Dooley, Yue Yang, and Xuguang Wang
What is already known:
- Turbulent mixing in the ocean surface boundary layer (OSBL) affects air-sea interactions, and the intensification of tropical cyclones (TCs).
- The critical Richardson number (Ric) dictates mixing but is often fixed in models.
- The HAFS-MOM6 is a coupled hurricane forecast model but is limited from its lack of ensemble spread and underrepresented uncertainty.
- To forecast TCs accurately, coupled systems require ensemble diversity and well-sampled background error covariances.
What this study adds:
- Perturbing Ric within the ensemble forecasts shows location dependent sensitivity in the OSBL.
- Observations from gliders, saildrones, dropsondes, and AXBTs were used to verify the perturbed experiment.
- Ric perturbations showed location-dependent differences in the atmosphere, air-sea interface, and upper ocean region, improving the ensemble spread and diversity.
- Incorporating Ric perturbations into the data assimilation system should improve uncertainty and forecast accuracy.
Abstract:
Turbulent mixing is a process in the ocean surface boundary layer that affects the air-sea interactions, which plays a key role in the intensification of tropical cyclones (TCs). To improve the numerical forecasting of TC intensification, accurately modeling the process of turbulent mixing is essential. However, uncertainties exist in the turbulence parameterization within the vertical mixing schemes, which can cause forecast errors, including both systematic biases and variability mismatches between forecasts and observations. Given the impact of background-error covariances on data assimilation and background ensemble under-dispersion at the air-sea interface, ocean physics errors may contribute to insufficient sampling. As a first step, this study perturbs the critical Richardson number (Ric) in the K-profile parameterization and uses novel observations from Hurricane Fiona (2022) to verify the background ensemble. Using the Hurricane Analysis and Forecast system coupled with Modular Ocean Model v6 (HAFS-MOM6), two sets of 40-member background ensemble without and with perturbing Ric are generated and compared. Verification against the novel observations showed that perturbing Ric increased ensemble spread and diversity in the atmosphere, air-sea interface, and ocean. This indicated an improved forecast uncertainty, and could be beneficial towards coupled data assimilation.