Axisymmetric Initialization of the Atmosphere and Ocean for Idealized Coupled Hurricane Simulations

• Published in: Journal of advances in modeling earth systems 2017-11, Vol.9 (7), p.2672-2695
• Main Authors: Aksoy, Altuğ, Zhang, Jun A., Klotz, Bradley W., Uhlhorn, Eric W., Cione, Joseph J.
• Format: Article
• Language: English
• Subjects: coupled simulations; Fields; hurricane simulations; Hurricanes; idealized simulations; initialization; Mathematical models; Mixed layer depth; Numerical models; Ocean models; Oceans; ocean‐atmosphere coupling; Perturbations; Regression models; Sea surface temperature; Simulation; Storms; Surface temperature; Upper ocean; Weather forecasting
• ISSN: 1942-2466; 1942-2466
• DOI: 10.1002/2017MS000977

A new vortex‐scale initialization scheme is presented for idealized coupled hurricane simulations. The atmospheric scheme involves construction of azimuthally averaged kinematic and thermodynamic initial fields based on historical composite data sets from hurricane reconnaissance aircraft. For ocean initialization, a statistical scheme is proposed to construct regression models among atmospheric and ocean fields in the hurricane inner core. For the numerical model, the Hurricane Weather Research and Forecasting (HWRF) model coupled with a one‐dimensional, diffusive ocean model is used with modifications to initialize with the observation‐based vortex and to ensure that the storm environment remains approximately steady. The primary goal in these simulations is to obtain steady state hurricanes of category‐1 intensity with characteristics typically observed during the hurricane season of the western Atlantic and Caribbean Sea regions. It is demonstrated that this is successfully achieved in the simulations. In an azimuthally averaged sense, regression models are found to capture about 70% of total variance for sea‐surface temperature cooling and up to 55% of total variance for mixed‐layer depth perturbation in the hurricane inner core. Furthermore, within the inner core of a hurricane vortex, it is found that storm speed contributes most to upper ocean perturbations, whereas characteristics of the atmospheric vortex contribute very little. The importance of storm speed in controlling upper ocean perturbations is strongest near the storm center, diminishing gradually toward no measurable impact beyond the immediate inner core. Key Points A new vortex‐scale initialization scheme is developed for idealized coupled hurricane simulations For the atmosphere, composite hurricane observations are used; for the ocean, regression models with atmospheric variables are utilized The importance of storm speed in controlling upper ocean perturbations in the hurricane inner core is demonstrated