Surface‐flux‐induced tropical cyclogenesis within an axisymmetric atmospheric balanced model

Abstract

AbstractThe WISHE (wind‐induced surface heat exchange) instability theory for tropical atmospheric disturbances suggests that maximum warming occurs at the position of maximum surface wind. Applying this concept to axisymmetric cyclones leads to outward migration of the radius of maximum wind without amplification. Therefore, the fundamental WISHE feedback loop cannot explain tropical cyclogenesis on its own. Additional restrictions for the surface‐flux‐induced heating are necessary.This study investigates in detail the effect of such restrictions on surface‐flux‐induced tropical cyclogenesis within the framework of a simple axisymmetric balanced model previously developed by the author. This model is based on the assumption of zero potential vorticity and formulated in potential‐radius coordinates. Heating is restricted by an efficiency parameter β that was previously introduced by Emanuel in a similar but more simplified model to account for the reduction of net latent heating in an unsaturated atmosphere. As in Emanuel's model it is found in the linearized and in the nonlinear model that a negative gradient of the efficiency parameter is necessary to initiate tropical cyclogenesis. In this case, wind amplification primarily takes place in the region where the gradient of the efficiency parameter is negative. This region migrates inward, shrinks due to frontogenesis and becomes the tropical cyclone's eyewall. It is shown that cyclogenesis becomes possible even when the surface heat exchange is wind‐speed independent. In the case of a zero β‐gradient outward propagation with no amplification takes place. If a cold anomaly is present at the vortex centre, the wind speed increases slightly but no cyclogenesis occurs. If on the other hand moisture fluxes can modify the initially uniform β‐profile, cyclogenesis appears after anegative β‐gradient is established. Horizontal diffusion is necessary to prevent a frontal collapse during the development and acts to transport momentum to regions where no wind is generated by heating. However, inward momentum diffusion does not accelerate the development as it was found previously in the model of Emanuel. This cyclogenesis process can already be qualitatively reproduced analytically by the linearized version of the model which additionally includes horizontal diffusion effects. Copyright © 2006 Royal Meteorological Society