Mechanism of concentric eyewall replacement cycle and associated intensity change

Concentric eyewall replacement is a common process for intense tropical cyclones (TCs). Prediction of its occurrence, evolution and associated intensity fluctuation has been a great challenge. To understand the mechanisms responsible for the replacement cycle and intensity change, two experiments were conducted using the Weather Research and Forecast model, in which the initial conditions and model parameters were identical except that the concentration of ice particle is enhanced in the sensitivity experiment. The secondary eyewall forms at a larger radius, the eyewall replacement process is slower and intensity fluctuation is larger in the sensitivity run than in its counterpart. The enhanced concentration of ice particles at the upper troposphere outflow layer produces a clear subsidence region (moat) surrounding the primary eyewall due to the reinforced cooling of ice particle melting as they fall into warm air. As a result of the presence of the moat, the secondary eyewall forms at a relatively large radius. By performing axisymmetric equivalent potential temperature budget, we found that the demise of the inner eyewall is primarily due to interception of the boundary layer inflow supply of entropy by the outer convective ring, whereas the indirect process that low entropy air is advected from middle levels to boundary inflow layers by downdrafts in the moat is not essential. The “interception” mechanism becomes less efficient when the secondary eyewall is located at a larger radius. Accordingly, the eyewall replacement is slower. After the dissipation of the inner eyewall, the outer eyewall has to maintain a warm core not only in the previous eye, but also in the moat. The presence of lower equivalent potential temperature air in the moat results in more significant weakening of the storm intensity during the eyewall replacement. The results here suggest that ice-phase microphysics is important for concentric eyewall replacement cycles, and monitoring the features of the moat region may provide a clue for prediction of the TC intensity change.