A novel statistical technique called space-time empirical normal mode (ST-ENM) is applied in a diagnostic study of the genesis of a secondary eyewall in a simulated hurricane using the nonhydrostatic, high-resolution fifth-generation Pennsylvania State University (PSU)-National Center for Atmospheric Research (NCAR) Mesoscale Model (MM5). The bases obtained from the ST-ENM technique are nonstationary, dynamically relevant, and orthogonal in the sense of wave activity.The wave activity spectra of the wavenumber-1 anomalies show that the leading modes (1-6) exhibit mainly characteristics of vortex Rossby waves (VRWs). These modes together explain about 75% of the total wavenumber-1 variance in a period of 24 h.Analysis of the Eliassen-Palm (EP) flux and its time-mean divergence corresponding to the total contribution from these modes indicated that in the lower troposphere VRWs not only propagate inward (outward) in the primary eyewall region where the radial gradient of the basic-state potential vorticity is large and positive (large and negative), but there is also wave activity propagating outside the primary eyewall. Consequently, maximum cyclonic eddy angular momentum is transported not only inside the radius of maximum wind (RMW) by VRWs in the primary eyewall region, but also close to the location where the secondary eyewall forms by VRWs propagating outside the inner eyewall.The fact that the critical radius for some of the ST-ENMs is contained inside the region where the secondary eyewall forms and the existence of a signal of maximum eddy cyclonic angular momentum flux propagating outward up to the critical radius suggests that a wave-mean flow interaction mechanism and redistribution of angular momentum may be suitable to explain important dynamical aspects of concentric eyewall genesis. FIG. 4. Axisymmetric tangential wind (m s 21 ) at 6, 12, 18, and 24 h. FIG. 5. Wave activity spectra of the (top) total pseudomomentum J (kg K 21 s 21 ) and the pseudoenergy A (kg K 21 s 22 ) and (bottom) individual terms of J and A of wavenumber-1 azimuthal disturbances. 466

No abstract

No abstract

No abstract

No abstract