Microphysical-Dynamical Interaction in Tropical Cyclone Intensification

Abstract

Producing timely and accurate tropical cyclone (TC) intensity forecasts remains one of the most difficult challenges facing meteorologists today. The state-of-the-art three-dimensional (3D) full physics operational models, in particular, have problems in simulating rapid intensification (RI), a situation where a TC intensifies dramatically in a short period of time. For example, Hurricanes Patricia (2015) and Maria (2017) increased their maximum sustained winds by 90 knots and 70 knots within 24 hours, respectively. The major objectives of this dissertation are to (a) explore the underlying reasons why the Hurricane Weather Research and Forecasting (HWRF) system, one of the operational models used at the Environmental Modeling Center (EMC), NOAA for TC prediction, often fails to predict RI of major hurricanes; (b) improve turbulent mixing and hydrometeor sedimentation parameterizations in HWRF so that the model can improve the capture of the dynamical-microphysical interaction in the TC inn-core region; and (c) advance our understanding of the internal dynamics that governs the RI of TCs. Our investigations show that HWRF is unable to generate appropriate sub-grid-scale (SGS) turbulent forcing above the boundary layer in the eyewall of a TC as a result of the lack of consideration of the cloud induced buoyancy. Incorporating a saturated Brunt-Väisälä frequency (BVF) in static stability calculation allows HWRF’s turbulent mixing scheme to successfully generate the buoyancy production of turbulence in the eyewall, which notably improves HWRF’s skill in predicting RI of TCs. Our analyses show that the HWRF microphysics scheme fails to produce hydrometeors with fall speed smaller than 0.2 ms-1, which are abundant in TC clouds according to in-situ aircraft measurements. The failure stems from the fact that the hydrometeor fall speed parameterization used in HWRF was developed in quiescent conditions that neglects the strong impact of convective currents on falling hydrometeor particles. To fully consider the impact of dynamic-microphysical interaction on hydrometeor sedimentation in the eyewall, a new parameterization of particle fall speed in non-quiescent conditions is developed and implemented in HWRF. The new scheme successfully generated the lightest hydrometeors that the operational HWRF fails to produce and allows the HWRF to realistically simulate the convection in the TC inner-core region. Using the model output, this dissertation clarifies the role of eyewall convection in modulating the outflow temperature of a TC and in triggering the positive feedback mechanism among surface winds, evaporation, and storm intensification underlying the RI of TCs.