Idealized hurricanes are studied at four different sea surface temperatures (SSTs) of 26°, 27°, 28°, and 29°C in six nested domains down to 62‐m grid size. For the maximum and the distribution of wind speed, domain D6 (62 m) is similar to domain D5 (185 m) showing convergence near 200 m. For SST of 26°, domain D5 does not have small‐scale turbulence structures like the other three cases, which is an extension of the previous work of Rotunno et al. (2009; https://doi.org/10.1175/2009BAMS2884.1) who only obtained resolved gusts at 62 m. The distribution function of speed changes between the 27° and 28° cases, the latter one having a second peak, related to a structural change. Finally, it is suggested that the damage potential is not simply determined by the maximum wind speed because of the distribution change so a new damage indicator that represents potential damage is proposed.

No abstract

No abstract

 The hurricane structure changes as sea surface temperature increases: the distribution function of wind speed changes from single peak to bimodal.  The scale of turbulent eddies increases as sea surface temperature increases. For hurricane intensity, a horizontal grid size near 200 m is sufficient for a converged solution.  A new damage indicator is proposed for hurricane disaster risks.

Wall model large eddy simulations (WMLES) are carried out to investigate the amplitude modulation exerted on near-wall small-scale motions by outer layer large-scale motions in the atmospheric boundary layer at high Reynolds number O(106–107). The properties of the mean and fluctuating velocities show good agreement with those found in previous studies. Furthermore, the positions at which there is no amplitude modulation found in the present study are consistent with those found in previous studies. A new phenomenon is observed, namely, that the value of the negative maximum correlation at high Reynolds number is smaller than that at low and moderate Reynolds number. Further investigation shows that the negative maximum correlation decreases with increasing Reynolds number, which could be explained by intermittency effects. There is good agreement of the correlation for different values of the Reynolds number when scaled with outer variables, which confirms that the large boundary-layer-height-scaled events that inhabit the logarithmic region are the source of amplitude modulation. This is confirmed by the locations of other characteristic points, which are independent of Reynolds number. In contrast, when scaled with inner variables, these characteristic points have a strong linear dependence on Reynolds number. Furthermore, the reversal in sign of the correlation corresponds to the crossover points of small-scale turbulent intensity and the local peak in the energy distribution, which gives the first and secondary crossover points a specific physical meaning. Finally, we provide an overview of the energy distribution, which gives an intuitive view of the outer peak energy site.